GITEX ASIA 2025 will bring together 700+ tech companies, featuring 400+ startups and digital promotion agencies, and 250+ global investors & VCs from 60+ countries.

The event will serve as a bridge between the Eastern and Western technology ecosystems and feature 180+ hours of expert insights from 220 global thought leaders.

GITEX ASIA 2025 is set to foster cross-border collaboration, investment, and innovation, connecting global tech enterprises, unicorn founders, policymakers, SMEs, and academia to shape the future of digital transformation in Asia.

GITEX ASIA 2025 will comprise of five co-located events:

• AI EVERTYTHING SINGAPORE – the AI showcase.

• NORTHSTAR ASIA – for startups and investors

• GITEX CYBER VALLEY ASIA – helping create a defence ecosystem for governments and businesses

• GITEX QUANTUM QUANTUM EXPO ASIA – Asia’s quantum frontier

• GITEX DIGI HEALTH & BIOTECH SINGAPORE – the healthcare revolution

GITEX ASIA 2025 will host a lineup of conferences and summits, exploring a range of transformative trends in technology and investment. Key themes will include AI, cloud & connectivity, cybersecurity, quantum, health tech & biotech, green tech & smart cities, startups & investors, and SMEs.

Sessions will include Asia Digital AI Economy, AI Everything: AI Adoption & Commercialisation, Cybersecurity: AI-Enabled Cybersecurity & Critical Infrastructure, Digital Health, and the Supernova Pitch Competition.

The event will bring together leading voices and ideas from different industries, including public services, retail, finance, education, health, and manufacturing.

Be part of the action at GITEX ASIA 2025 and witness the future of technology unfold in Singapore. For more information and updates on GITEX ASIA, visit www.gitexasia.com

Social media links: LinkedIn | X | Facebook | Instagram | YouTube

The post GITEX GLOBAL in Asia: the largest tech show in the world appeared first on TechWire Asia.

The buzz surrounding AI has multiplied over the past three years, with headlines about rapid technological advancement, adoption, business transformation, job market effects, ethical quandaries, and the impending arrival of artificial general intelligence all highly visible.

As of the beginning of 2025, we’ve already seen major moves from the US, including the Stargate Project, a $500 billion investment in AI infrastructure from SoftBank, OpenAI, Oracle, and MGX. Meanwhile, DeepSeek’s emergence from China as a sophisticated cost-effective reasoning model has shaken markets.

Between all the headlines and hot-takes, a clear message is also being pushed to businesses: adopt AI or risk being left behind.

For retailers, this is creating real pressure to act. Customer expectations are shifting towards personalised shopping experiences that AI can enable. Meanwhile, businesses feel they must keep pace with technological changes to maintain market share, while AI also promises opportunities to improve operations and increase revenue.

The opportunity in Southeast Asia

Southeast Asia is a hot market for AI adoption. With competition among major players heating up and technology providers integrating AI into their offerings, AI is now more accessible to mid-market retailers as well as major enterprises. The region is also well-suited to AI integration, as the market typically moves quickly to implement new technologies once proven.

While affordable AI technologies and the appetite to harness them exist among retailers, there is a key challenge often absent from the online discourse or in the traditional press: determining specific use cases that add value.

From chatbots to retrieval augmented generation systems (RAG) and autonomous AI agents, organisations are still working out how to implement AI effectively and efficiently. The excitement around AI’s potential is clear, but businesses need to ensure any implementation delivers tangible value, rather than just existing for the sake of it.

This leaves retailers in a challenging position. The pressure to adopt AI is significant and technologies are maturing fast. Companies have to balance pressure and the identity of practical applications that create real value for their business and customers.

What AI adoption looks Like

There are many ways a retailer could potentially integrate AI applications into its operations, depending on size and needs.

For example, a retailer might sign up for one of the many generative AI platforms available to produce promotional images and graphics for marketing purposes.

If the company operates an online presence, another entry point might be an LLM-driven chatbot that can handle customer inquiries about things like opening hours, item availability, specials, and return policies.

One increasingly popular and proven way that retailers in Southeast Asia can integrate AI into their operations is through solutions that use customer data to improve the shopping experience. By another name, this is authentic personalisation.

Personalisation makes a retail store so much more than just another store. It allows brands to provide a shopping experience that is tailored to each customer’s individual preferences.

Consider the following example. In the future, before setting foot in the store, customers of Cold Storage in Singapore might receive completely personalised product recommendations that consider not only their past purchases, dietary preferences and lifestyle, but also real-time contextual factors like the weather or local festivals.

The recommendations could be delivered through an app, via email, or social media. The app could also create shopping lists based on weekly patterns and trends, or upcoming holidays, helping customers save time and ensure they don’t forget any regular staples.

Another key pre-shopping AI opportunity is personalised offers and promotions, including challenge offers. Challenge offers provide a gamified experience where customers are increasingly rewarded for meeting specific targets, like spending a certain amount in a set period. The challenges can be tailored to a customer’s preferences, presenting goals or targets for product groups they like or buy often.

Personalisation also extends to the in-store experience, where recommendations might pop up in the app based on where customers are in the store. Customers might also scan products with their phones to receive reviews and recipes they might like. Taking this a step further, supplier-funded personalised ads for attractive items could also be generated on a customer-by-customer basis.

A proven way to bring AI to business

Personalisation and gamification solutions for retail can help retail brands increase customer satisfaction and loyalty. Retailers don’t even need to have an existing loyalty programme to get started.

In the case of Eagle Eye’s offering, for example, a retailer in Southeast Asia could deliver personalised challenge promotions, powered by AI, in as little as five weeks. This represents a speed to market that matches the region’s hunger to roll out technology solutions quickly.

Such solutions have already been delivered in other markets to great effect. For example, in the United Kingdom, major grocery brand Tesco has adopted AI and is using it to bring benefits to its customers.

Tesco launched Clubcard Challenges in May 2024. The is a loyalty-integrated gamification initiative that utilises AI to create customised, shopper-specific challenges.

Clubcard members are invited to participate in the game, and are served 20 distinct challenges, like spending £20 on summer BBQ supplies for a chance to collect up to £50 in Clubcard points. Once all tasks are completed, they can win additional rewards.

In other markets, major coffee chain Starbucks uses its Deep Brew technology to analyse customer preferences and contextual data, enabling personalised recommendations like suggesting cold drinks to specific customers during warm weather.

Similarly, French supermarket chain Carrefour has partnered with Eagle Eye to gamify its MyClub loyalty program, creating customised challenges and goals based on individual shopping patterns and purchase history data.

Make it happen

Examples and real-world case studies such as those above demonstrate how retailers in the region can create powerful customer experiences, drive loyalty, and increase profitability without extensive lead times or long implementation timelines. Rollouts can be quick and remain cautious. Pilot programmes can be run to test effectiveness before moving to full-scale adoption.

Taking the first steps in AI-driven personalisation with a partner like Eagle Eye means retailers in Southeast Asia can start with innovative solutions like challenge offers quickly and easily, taking the anxiety out of being left behind in the AI race and joining other early-adopting global brands.

(Image source: Unsplash)

The post How retailers in Southeast Asia can adopt AI quickly appeared first on TechWire Asia.

Many organisations are employing a shift-left approach to cloud security to address these challenges. They include security early in the development lifecycle so teams can identify vulnerabilities before they become too severe. By baking security into the development process, businesses lower risk and build confidence in their cloud-based operations.

Before diving into the Shift-Left approach, it’s necessary to understand why cloud application security is so important.

Why cloud application security matters?

With organisations increasingly looking toward cloud services for storing and processing sensitive information, the associated risks must be acknowledged and implemented appropriately in cloud application security testing. This would ascertain and improve the security features within the cloud environment while protecting critical data and ensuring consumer confidence. This assessment includes inspecting applications, databases, networks, and possible vulnerabilities.

Organisations can utilise different testing methods to identify areas vulnerable to breaches or cyberattacks, thus allowing them to enforce necessary measures to intensify their defences against cybercrime.

Cloud security is a shared obligation of the consumer and the cloud provider. The shared responsibility notion identifies three types of responsibilities. The cloud provider is always responsible for the security of its infrastructure and networks. In contrast, the customer is responsible for identity and access management, encryption and security of cloud-based data assets, and compliance. The last category includes responsibilities that differ depending on the service model: Software as a Service (SaaS), Infrastructure as a Service (IaaS), or Platform as a Service (PaaS).

Cloud application security testing software uses industry-leading approaches (SAST, DAST, IAST, and SCA) to discover the most prevalent security vulnerabilities, from online to mobile to open-source. Application security tools prevent risks in applications before they get into production.

The shift-left approach: Securing from the start

Shift-left security includes integrating security at the earliest stages across the software development lifecycle. This provides application security testing in the planning and development phases to preempt security demands as soon as possible and mitigate potential problems later in development. Shift left indicates a left movement, facilitating collaboration between development, information security, and operations.

Shift-left for cloud security minimises the risk of application security vulnerabilities surfacing in production while streamlining time-to-market. Here are some key benefits of adopting shift-left for cloud security:

• Early vulnerability elimination: By incorporating security into the SDLC’s early phases, developers may find and remedy vulnerabilities before they become more complex and expensive to resolve.
• Improved collaboration: Shifting left fosters early engagement between security and development teams, which improves communication and alignment around security goals.
• Enhanced security posture: Dealing with security problems early in SDLC might reduce the possibility of data breaches, cyberattacks, and compliance violations.
• Lowered costs: Addressing security vulnerabilities before the development stage is usually less expensive than dealing with them later in production.

Organisations are looking for ways to integrate security into the development process while empowering developers to create secure and reliable solutions—without requiring them to become security experts or slowing down application development. Shift-left security helps achieve this by significantly alleviating security concerns associated with cloud-native software and application development. As the cloud continues to dominate digital transformation, shifting left will become a critical security best practice for the 85% of organisations projected to adopt a cloud-first strategy by 2025.

Key features of enterprise-grade cloud application security software

Comprehensive testing capabilities

To ensure robust protection, enterprise-grade solutions provide multiple layers of testing:

• Dynamic application security testing (DAST): DAST tools look for web application vulnerabilities while running. They allow users to scan running apps and APIs for potential vulnerabilities in the production environment.
• Static application security testing (SAST): It scans source code in applications and APIs for vulnerabilities in the early stages of development. SAST offers scanning solutions for developers, supporting diverse technology landscapes.
• Interactive application security testing (IAST): It monitors apps and APIs to detect and resolve issues without slowing development. It unites SAST and DAST by analysing the application during runtime with deeper code insights, providing more actionable results.
• Software composition analysis (SCA): It covers apps from vulnerabilities caused by open-source software components. SCA integrates into multiple stages of the application’s lifecycle to quickly assess components within specified folders, containers, or images.

Automated scanning integrated with DevSecOps workflows

Modern security testing software integrates build environments, DevOps tools, and integrated development environments (IDEs) to create a frictionless application security testing experience. As organisations increasingly embrace automation in their DevOps workflows, DevSecOps emerges as a crucial approach, embedding security controls directly within the continuous integration and delivery process

By incorporating automated security measures into the production cycle, organisations can address vulnerabilities before applications go live, minimising inefficiencies and mitigating risks. This integration ensures that security is no longer an afterthought but a critical component of the development process. DevSecOps facilitates this transition by automating cybersecurity and managing the CI/CD toolchain. This approach enhances application security while simplifying the development lifecycle.

AI-driven insights for prioritising vulnerabilities

The evolving cyber threat landscape and surge in connected devices have drastically expanded organisations’ attack surfaces. Managing vast networks, countless attack vectors, a talent shortage, and overwhelming data volumes has become daunting. AI and ML have emerged as vital tools to address these challenges. By automating threat detection and response, AI can process data at a scale and speed surpassing human capabilities.

With the continued development of cloud services, AI is pivotal in automating tasks such as:

• User access management: Reducing the risk of unauthorised access through intelligent automation.
• Error mitigation: Minimising human error in security configurations and processes.

Organisations can improve their cybersecurity measures and optimise operations by using AI-driven methods to prioritise vulnerabilities.

Conclusion

In the ever-changing threat landscape, proactive security is more than an option; it is essential. Approaches like shift-left testing are crucial for detecting vulnerabilities early on, strengthening your security posture, and allowing continuous innovation during your digital transformation. Schedule an application security demo today and take the next step toward protecting your business. Discover how application security on the cloud may provide you with sophisticated capabilities and strategies for staying ahead of threats. Explore its potential today and rethink your approach to security.

The post Cloud application security: The importance of shift-left testing in development appeared first on TechWire Asia.

This article explores how these models differ and provides practical guidance to help you make an informed decision for your fine-tuning needs. Read on to understand how to align your goals with the right model and optimise your results.

Overview of open-source and proprietary models

Fine-tuning large language models (LLMs) requires careful consideration of the model type. Open-source and proprietary models both offer unique advantages, but they come with specific challenges. Understanding these differences is vital for selecting the right approach for LLM fine tuning.

The rapid growth in LLM development highlights the need for efficiency. Over the last three years, computation used for training LLMs has surged by over 574,000%. This makes focusing on the most relevant data and model type important in optimising fine-tuning.

Open-source models

Open-source models are freely accessible and highly customisable, making them a preferred choice for many.

Proprietary models

Proprietary models are developed and maintained by private organisations. They are often optimised for specific use cases, offering strong support and advanced capabilities.

Thus, both types of models can suit different needs, depending on factors like budget, technical capabilities, and the level of customisation required. By aligning the choice of model with project requirements and focusing on data relevance, organisations can ensure successful LLM fine tuning. The decision also impacts the efficiency of annotation processes, further boosting fine-tuning outcomes.

Advantages and challenges of each model type

Selecting the right model for LLM fine tuning often depends on balancing the strengths and limitations of open-source and proprietary options. Each type brings specific benefits and challenges to the table, which can shape how well they meet the needs of fine-tuning tasks.

Open-source models: Flexibility with trade-offs

Open-source models are popular for their adaptability and cost-effectiveness. Developers have the freedom to modify them, which helps create tailored solutions for unique uses. However, this flexibility comes with certain challenges.

Advantages:

• Customisability: Open-source models allow changes to their architecture or training processes, making them suitable for specialised tasks,
• No licensing fees: Using open-source models eliminates the cost of access, making them a budget-friendly option,
• Community support: Many open-source models benefit from active communities, providing access to shared knowledge and updates.

Despite these advantages, some issues may arise. They demand more technical expertise, as deploying and fine-tuning them requires deeper knowledge. And, they might lack the extensive datasets and optimisations seen in proprietary solutions.

Proprietary models: Convenience at a cost

Proprietary models focus on providing ready-to-use solutions with professional support. Their performance often stands out, but their limitations can affect long-term use.

Advantages:

• High performance: Proprietary models are typically pre-trained on diverse and extensive datasets, which ensures strong baseline accuracy,
• Ease of use: The models often come with user-friendly tools and documentation, reducing the complexity of implementation,
• Support and maintenance: Developers can rely on dedicated customer support for troubleshooting and updates.

However, the cost of licensing can be prohibitive, especially for small-scale projects. Proprietary models also restrict modifications, which may limit customisation options.

Deciding Between the Two

The choice between open-source and proprietary models often depends on project-specific needs. For example, open-source models might suit research projects with limited budgets and high customisation needs. Meanwhile, proprietary models can work well for enterprises prioritising performance and quick deployment.

By carefully weighing these advantages and challenges, teams can align their model choice with their fine-tuning goals. Moreover, ensuring high-quality annotated data remains critical in maximising the performance of both model types.

A guide to choosing the right model for fine-tuning

Selecting the right model for fine-tuning involves understanding your project’s goals, resources, and constraints. Both open-source and proprietary models have their strengths, so the decision depends on aligning these with your needs. The guide outlines practical steps to simplify the process.

Step 1: Define your goals

Start by identifying the purpose of fine-tuning. Determine whether the model will handle general or specialised tasks. For highly specific use cases, open-source models can offer greater customisation. Proprietary models are better for tasks requiring robust, out-of-the-box performance.

Step 2: Evaluate technical expertise

Assess your team’s technical skills. Open-source models may require advanced expertise for setup and fine-tuning. On the other hand, proprietary models often include tools and support, making them easier to manage for teams with limited technical experience.

Step 3: Consider budget constraints

Review your budget to decide whether the licensing costs of proprietary models are manageable. Open-source models are free but may require additional investments in infrastructure or expert developers.

Step 4: Focus on data quality

High-quality annotated data is critical for achieving fine-tuning success, regardless of the model type. Collaborating with professionals offering data collection services ensures accuracy and consistency, reducing errors during training.

Step 5: Weigh long-term goals

Think about long-term needs. Open-source models provide greater flexibility for evolving tasks, while proprietary models may require ongoing licensing commitments.

By following these steps, teams can make informed decisions that match their resources and objectives. Choosing the right model type is not just about performance – it’s about ensuring the model fits your fine-tuning workflow and produces reliable outcomes.

Final thoughts

Choosing between open-source and proprietary models for LLM fine tuning requires a clear understanding of your project’s priorities, resources, and long-term goals. Each model type has unique strengths and challenges, making careful evaluation essential.

By aligning model selection with your specific requirements and focusing on high-quality data annotation, you can enhance fine-tuning efficiency and achieve better results. As advancements in AI continue, staying adaptable will help you make the most of evolving technologies.

The post Comparing open-source and proprietary models in fine tuning appeared first on TechWire Asia.

Remove old network records

Using a Mac cleanup tool (such as the one available via this link) to delete older networks can be very effective. Similar apps will have different methods, and not all will automatically remove old networks during a software cleanup, requiring the user to visit a dedicated menu to find the feature. For a lot of people, old networks can lead to problems if left in Network Settings, especially if you rely on Wi-Fi.

Delete and then re-add your Wi-Fi connection

Go to System Preferences, Network settings and remove the current Wi-Fi connection by highlighting it and pressing the “-” button. Then, re-add the necessary details by selecting Wi-Fi from the drop-down menu, press Create, add your network again, then Apply.

Flushing the DNS cache

As more info accumulates on sites and properties your computer connects to, you can end up with a DNS cache with out-of-date entries. These can lead to issues with your device and flushing the DNS cache is a way to ensure the cache starts to build with up-to-date data.

Flushing the DNS cache is something you can do with a Mac cleanup tool or manually, and it can be a very quick, seamless operation.

Reset the SMC

The system management controller (SMC) controls low level functions. It controls the computer’s fans, handles power management, battery life and so on. Apple has SMC reset guidelines you can follow.

PRAM reset

If the SMC reset doesn’t do the job, a PRAM reset might be what you need. The parameter random access memory, which is maintained even when the Mac is off, holds the computer’s core configuration data. Resetting it may help as it resets networking info too. Obviously, that’s not something you want to mess around with unless you know 100% that other methods didn’t work.

The simplest way to reset PRAM is to shut down your Mac, press the power button and then hold down Command+Option+P+R. Once you hear a second chime, let go of the keys. The entire process is very fast and effective.

However, a PRAM reset is something that you should try last, if the other methods did not have the desired effect. What matters is to ensure the problem is tackled one step at a time.

Conclusion

These methods are effective if you have networking problems and want to solve them as fast as possible. Just removing and re-adding the network may not be enough, and you may need to do a SMC, PRAM reset or DNS cache flush. Follow each option in turn, and good luck.

The post Common ways to fix networking issues on a Mac appeared first on TechWire Asia.

Connecting with consumers in a meaningful way

Today’s brands, in and outside of Southeast Asia, face a constant challenge: Cutting through the noise and connecting with consumers in a meaningful way. Traditional marketing approaches – often focused on chasing fleeting trends or aligning with pre-planned events – are proving increasingly ineffective.

This approach moves beyond simply reacting to trends; it anticipates them. Instead of attempting to dictate what’s ‘in,’ successful brands in Southeast Asia are leveraging data analytics to understand the deeper currents shaping consumer behaviour. By identifying and responding to emerging interests and sentiments in real-time, successful brands can craft highly relevant and resonant messaging. The is not about creating trends, but reacting organically to them – capitalising on spontaneous moments and authentic connections.

How the Singapore Grand Prix got it right

A compelling example is the Singapore Grand Prix, where Singapore Airlines went beyond simple alignment. According to Singapore’s Ministry of Trade and Industry, since it began in 2008 the F1 Singapore Grand Prix has attracted more than 550,000 international visitors and generated around $2 billion (SGD) in incremental tourism spend.

By sponsoring and aligning strategically with event, the airline capitalised on increased passenger traffic and exposed its branding to millions of viewers tuning in globally. In understanding the broader cultural context, Singapore Airlines amplified its presence – highlighting the importance of understanding the wider culture of an event or trend to maximise impact and ROI.

Real-time analytics and contextual relevance

That said, the success of agile, data-driven marketing hinges on several key elements:

• Real-time data analysis: The ability to monitor and interpret real-time data is crucial. This allows brands to identify emerging trends, gauge audience sentiment, and adjust messaging accordingly. Tools providing real-time analytics on consumer interests, social media conversations and search trends are vital.
• Rapid response mechanisms: Once a trend or opportunity has been identified, brands must be able to act swiftly. This requires efficient internal processes, streamlined approval workflows and a culture of overall responsiveness.
• Contextual relevance: Messaging needs to be tailored to the specific context and culture of each target audience; generic, one-size-fits-all approaches are ineffective.
• Authentic storytelling: Consumers are increasingly discerning; they can easily spot inauthentic marketing attempts. Brands need to craft compelling narratives that resonate with their values and aspirations.

Data, tech and storytelling: the cornerstone of brand success

In the Southeast Asian marketplace, agility is the cornerstone of sustainable brand success. Data-driven insights, targeted media activation, and authentic storytelling are just a few tools that allow brands to move beyond reactive trend-chasing, and towards more proactive, meaningful engagement with their audiences. To stand out in this competitive landscape, the capacity to adapt and react to the ever-changing preferences of consumers will determine who thrives here… and who gets left behind.

The post Brands: Are you chasing trends or predicting them with tech? appeared first on TechWire Asia.

In this regard, contextual content services can change the workplace drastically. They help organise work schedules and tasks, reduce the amount of information people need to deal with, and ensure that workers can focus on their primary tasks without being distracted by redundant information.

Employees using content enhancement tools can derive insights faster, work together better, and complete more tasks.

Contextual content elevates operational content service approaches by delivering the exact information it’s needed based on job or role. Let’s say a seller needs information about a customer. With contextual content management, they don’t have to look through files or systems to find what they need; they get it right when needed. Contextual content services give each user and present the correct information at the right time.

The issue of burgeoning data

The average worker spends 30% of their day looking for the right information. Most professionals get overwhelmed by a new influx of information and systems usually suffer from information overload. Employees often waste time reading emails, papers, or data that need to be relevant in the context of their current task.

Contextual content management tools can help with these problems with intelligent document processing (IDP). The systems extract insights from diverse documents to elicit smarter responses at enterprise scale, where billions of documents need to be processed simultaneously.

IDP can help customers, employees, and partners access information and documents anytime, anywhere. Employees can focus better, work faster, and stay efficient by finding accurate information at any given point in time.

Why would you need contextual content management?

Contextual content is effective in many ways:

1. Work is done faster and better

People won’t waste time looking for information when they can get it quickly. Contextual content services immediately fetche documents, data, or contact information required for the completion of a task. This speeds up the decision-making process.

For example, a project manager working on the launch of a new product can get reports, customer feedback, or marketing plans right away without having to look through materials that aren’t relevant: they don’t have to do as many administrative chores and can focus on the bigger picture.

2. Better collaboration as a team
   

Collaboration is a key to success in any enterprise. ECM solutions ensure effective collation of documents and data in one repository and help authorised employees automate processes across teams.

Everyone on a design team, content writers and planners, can use the same set of up-to-date information to fuel their productivity. This ensures everyone is on the same page, reduces misunderstandings, and speeds up work.

3. Unique experience for each user based on context of operations

Contextual content management services ensure that each person receives tailored information essential to routine tasks. Whether you’re a business leader wanting an overview in the form of a presentation or a customer service professional requiring details about a product, you get content customised to your needs.

4. Real-time intelligent data extraction helps you make better choices

You need to know the right things at the right time to make good choices. Contextual content management offers insights and analytics that help employees make choices quickly, based on real-time data.

For instance, employees can get real-time information about how well they’re doing in sales or get quick customer feedback, letting them act quickly and confidently. Using AI-powered capabilities enables quick and accurate content extraction, automatic segregation of documents, and powerful sentiment analysis.

Takeaways

In the enterprise world, data powers operations. One can enable anytime-anywhere access with such systems while powering real-time collaboration among stakeholders. Contextual Content Services Platforms drive a hybrid culture based on data integration with multiple systems. Systems can be used to empower employees with real-time smart searches and handy recommendations that make workflows smoother and operations seamless and streamlined.

(Image source)

The post Boost employee productivity with intelligent, contextual content management appeared first on TechWire Asia.

One business function that has been working with machine learning in operations is customer support and/or call centres. Often in the form of chatbots or smart assistants, we use them every day to interact with retail companies, utilities, government offices and so on.

The Business-Focused AI Use Case

But as AI advances, use cases where organisations can leverage the technology are growing. To explore these, we turn to organisations that are early adopters of AI in customer care. These businesses are further down the AI path than many others. Tech Wire Asia had a conversation with two executives whose organisations are at the forefront of deploying AI in contact centres and, more broadly, across industries.

Dave Flanagan is Head of Digital and Conversational AI at Nexon Asia Pacific, a digital consulting and managed services organisation in Australia that assists its customers to implement digital transformation roadmaps, identifying key areas where AI can add the most value and tailoring solutions that integrate seamlessly with their existing systems. Phillip Townsend is the Strategic Director of Innovation, APAC, for Genesys, the market leader in AI-powered experience orchestration, delivering seamless and personalised customer and employee experiences.

Employee Experiences

In addition to the intelligent assistants already in use by many customer care functions, AI brings significant value in improving employee experiences. By giving contact centre staff access to AI-powered tools, job satisfaction increases, and the daily workload becomes more efficient. Improving staff conditions lowers staff churn rates, creates clear career progression opportunities, and – a key metric in the boardroom – makes for more loyal customers who get better service.

Phill Townsend described how early machine learning algorithms were deployed. “The first thing that was identified as an opportunity for AI was to drive automation to reduce cost in the organisation. A lot of technology innovations can be deployed really poorly. The speech IVR of the early 2000s didn’t really serve a purpose for the customer, but was adopted to drive technology. Whereas now, there’s a belief in organisations to use tech to augment the agent. At Genesys, a lot of our R&D and our investment is going into leveraging augmentation in the agent’s processes, making sure it’s 100% right for our customers, and then shifting that into virtual agent capabilities.”

The Science of Compliance

There is a great deal of awareness in customer care functions of issues around data compliance, sovereignty, and security. That’s especially relevant given AI’s need for data to learn from and act on.

“When it comes to consolidating vast amounts of data into meaningful insights or making informed decisions backed by data, organisations must consider where their data will reside, ensuring it’s secure while leveraging AI tools to uncover actionable trends. This approach helps unlock new insights, empowering both the organisation and its staff to achieve better outcomes,” said Dave Flanagan of Nexon.

“Embedding compliance into AI development and practices means organisations must handle data management with a corporate-level perspective to prevent oversharing or unwarranted access to personally-identifiable information. By incorporating these practices, organisations can ensure that their data use aligns with regulatory requirements and safeguards sensitive information. As AI continues to be leveraged increasingly in call centres, it becomes even more crucial to prioritise compliance and robust data governance to maintain trust and uphold regulatory standards,” Dave said.

Expecting AI to independently decide upon a course of action, therefore, is one of the more high-risk projects that call centres might pursue. “I don’t know many organisations that do it well, because it involves multiple data sources such as customer history, customer profile and real time journey information. That’s high complexity, high risk. There are so many opportunities to be had with AI that are further down the risk scale yet bring immediate value,” he said.

The Smart Employee Assistant

This brings us back to using AI to augment the human worker and improve their overall experience. Phill pointed out that using AI to remove simple customer engagements from agents left them with the more stressful tasks. “[They were] dealing with far greater complexity in every single engagement. And they probably weren’t being supported well within the business.”

To improve employee experience, it is important to look to AI to provide support in everyday interactions.

“An AI-powered agent copilot can surface the right knowledge based on the conversation that customer is having in real time. That takes away stress from the agent. It’s like the bumper rails on a bowling alley. If the ball’s going down the middle, leave it alone. If we start to sway off track, let’s use some bumpers to guide the agent.”

The additional benefit of using software as an employee’s virtual assistant is that employee training and onboarding cycles get shorter. Phil said, “Instead of six and eight week training and onboarding cycles, you dramatically shrink timeframes by an AI surfacing the knowledge without the agent having to know everything.”

Additionally, because call handlers become fluent faster, staff can transition to other roles in the company, sometimes in other departments. This makes the candidate selection processes for open positions much simpler and less costly to the wider business.

Stepwise Implementation

And like any software rollout, the key to success is to proceed stepwise and address specific problem areas in the business. “Start small, ensure success,” Dave said. “Pick those right use cases that we know we’re going to get right, because we still need to build trust, not only with the customers, but with people inside the organisation.”

The internal buy-in for AI is particularly relevant because the media storm around AI means people can sometimes fear the technology. “It comes back to that trust. If you don’t achieve trust by success, you will encounter roadblocks pretty quickly,” he said.

To learn more about how artificial intelligence can enhance employee experience in the contact centre and maximise its value, consider partnering with Nexon, a company with extensive experience in delivering tangible results. To learn more about Nexon Asia Pacific’s practical AI implementations, visit here. To learn more about how AI-powered experience orchestration is delivering the future of customer and employee experiences, visit here.

The post Unlocking AI’s True Potential: Transforming Customer Care and Employee Experience appeared first on TechWire Asia.

In an exclusive interview with Tech Wire Asia, we asked the CEO of Affinidi, Glenn Gore, whether he thinks that commercial approaches by companies to data privacy are changing in accordance with public dissatisfaction with how their personal details are being used and lost.

“It’s evolving, and it’s unfortunately heading in the wrong direction, especially with the rise of AI. We know that last two years has seen incredible innovation from GenAI, but it is trained on public data sets. I think the world’s been a little shocked at how some of those public data sets have been used. You’ve seen that everywhere, from songwriters, musicians, script writers to social boards, where that data is being consumed at massive scale.”

Even with extensive customer records combined with available third-party records, there are issues with building customer profiles. The fragmentation of data means that companies can’t produce accurate pictures of any one individual regardless of whether they use AI algorithms or not. Building a profile of ‘the perfect customer’ and providing the goods and services they desire is impossible, because every entity has different personas depending on how they exhibit themselves online. A consumer might buy a certain brand and type of goods from one retailer, but will present quite a different digital profile when engaged in hobby activities online.

Affinidi’s Trust Network (ATN) provides a decentralised framework that prevents data misuse, such as hacked credentials and unauthorised access to personal details, while enabling secure and consensual data sharing. Through the use of Decentralised Identifiers (DIDs) and Verifiable Credentials (VCs), the ATN ensures that users have full control over the verified data they disclose to companies or third parties, so enhancing privacy and trust in digital interactions.

“Particularly for younger generation who’ve grown up in a digital world, transparency and trust, I think, become the new economy,” Glenn said. “I think companies that can establish trust with their customer base faster and better than others enter this flywheel effect, where the more trust they engender within their customer base, the more that customer base is actually going to be willing to share about themselves.”

The flip-side for the consumer is, Glenn explained, “The more I share about myself, I can see the experience changing, and I see value being created from that. I’m going to trust that [a company] is using my data for the right purpose, and I’m going to want to share more with them. And this is where companies that have that flywheel effect around trust are almost going to be untouchable.”

Holistic Identity and the Affinidi Iota Framework

The concept of ‘holistic identity’ is an integral part of the ATN. It is about giving individuals access and control of their own data. The Affinidi Iota Framework enables organisations to request only essential data points from individuals with their explicit consent, without organisations requiring to store non-essential details. Snippets like a person’s legal age or the professional certifications they have earned can be confirmed without passing on any additional details: there is a big difference between proving that a shopper is over 21 and providing full date-of-birth, for example.

Glenn uses the example of a bank loan, that at present requires a mass of compliance measures on the part of the bank, and the provision of highly sensitive information by the prospective lendee.

That’s two problems, he told us: the information supplied to the bank can be used for inference about the loan applicant, and the bank now has data that it has to protect.

An Iota (from the Affinidi Iota Framework) is essentially a query – a predefined set of rules or conditions that allows organisations to check whether certain criteria are met, without needing access to underlying data. It’s a privacy-preserving tool that allows computations to be made on personal data, while keeping both the data and the specific query details secure and confidential.

In this instance, if a bank provides an Iota that defines the criteria for loan eligibility, a prospective borrower can run this Iota query against their personal data without revealing any of their information to the bank. At the same time, the borrower remains unaware of the bank’s specific criteria. This creates a secure, privacy-preserving way to assess loan eligibility, where neither side gains more knowledge than necessary.

“Several powerful things happen in this process. The Iota query accesses only the relevant data stored in the individual’s personal vault, ensuring that no unnecessary information is leaked. The bank remains unaware of the applicant’s identity or personal details throughout the process. If the borrower doesn’t meet the criteria – for instance, due to insufficient savings – the Iota query returns a result indicating that the loan cannot be approved. At this point, the borrower has the option to share this outcome with the bank, but is under no obligation to do so, maintaining control over their private information.

“[Later,] they meet the criteria. ‘Do you give consent to share the result of this?’ And as part of this sharing of the result, the Iota will share just the input data, say, value of liquid assets and being paid at least $5,000 a month. But the loan applicant is not sharing pay slip data [which contains extraneous yet highly sensitive information]. Now a consent record goes back to the bank saying a verified identity, issued by a trusted party, of [customer name] was run against an Iota criteria, and they passed the checks. They can now choose to onboard as a customer.

“You can apply that to anything. Healthcare: Have you had blood work done? Dating sites and so on. You can use it for anything you want, and that’s really the power of it.”

The power of the message

A further difference in the ATN is the medium messages travel through. SMS and email were designed so that the owner of the account or number can be contacted by anyone. Affinidi Messaging is a decentralised messaging standard where the default position is you cannot send or receive messages.

“What’s the utility in that? Well, actually, that’s the first big safety feature of it. As you onboard, and you establish customer relationships with third parties, what you’re really doing is establishing a relationship between yourself as an identity and the third party. If I’ve opened a bank account with a bank, I give permission for it, with its known public key, to contact me. The message can be cryptographically signed and accessed only with my private key so I can be sure that this is absolutely the bank contacting me. So now I have a very secure communications channel at a relationship level between myself and the bank. Or between myself and the site I just bought some shoes from.”

So the Affinidi Trust Network provides privacy and encrypted communication with verified parties (which, Glenn told us, can be humans, businesses, AIs, machines or any digital system), and everything operates on the basis of trusted relationship and consent.

All about the CX relationship

But if we shelve the technical aspects of the framework, what we have is a way that trust can be manifest digitally: trust in a vendor, and trust in the truth of identity. It’s that basis that will form the next generation of CX, quite different from the scattergun approach to what we call customer experience at present.

“If you just strip all the fancy technology away, what we’re building is to allow companies or businesses to build trusted relationships with their customers, to allow their customers to share more about themselves, in order to get better experiences. And if you do that right, that creates a competitive advantage. The current technology stack does not allow that to occur, which is why we’re talking about a holistic identity approach. It touches everything from the concept of identity to authentication and authorisation as part of your onboarding, to how we communicate and share between each other using messaging, to consent management and privacy, with confidentiality at a personal level.”

Compared to the single sign on systems we use currently, Glenn says the ATN is “just a better way to manage identity and the relationships that identities create between themselves.”

But the network effect is difficult to create. Outside of Europe, India is one of the first countries that mandates consent from the individual for data sharing, and addresses the concept of a third-party data manager.

“I look at the DPDPA [enacted in 2023 – pdf] and say, that fits perfectly with what Affinidi is building because we can help facilitate the consent. This is where we’re starting to see the first real drivers. Because the great thing about DPDPA, is it’s government backed. Businesses in India must use these types of capabilities, so the planets are aligning. It’s still going to take time, but I think once they do start to align, I hope that there’ll be quite a rapid tipping point.”

The eyes of privacy advocates are firmly on Affinidi, while observing the effects of DPDPA as it takes hold. The eyes of every brand that wants to build more meaningful and trustworthy relationships with customers and third-parties should be on the same places.

To find out more about Affinidi, the Affinidi Trust Network and the Affinidi Iota Framework, head over to the relevant web pages.

The post The emergence of trust: next generation CX on next-gen tech appeared first on TechWire Asia.

The loyalty, promotions, and personalisation world is rapidly changing, and many organisations are contemplating how to best leverage exciting new technologies and opportunities that are arising in the space.

Deciding whether to build or buy a solution is a common conundrum. From time-to-market to hidden costs and lack of support, what follows are eight reasons technology buyers should deploy a ready-made, fit-for-purpose solution over trying to build it themselves.

The hidden costs of development time and resources

One of the biggest misconceptions people have when it comes to building a solution in-house is that it will be cheaper. The prospect of saving on licensing costs is attractive to businesses and this is often reflected by technology decision-makers.

However, going down the in-house route often means months – sometimes even years – of development time. As teams undertake a solution build, it’s not uncommon for companies to need to invest heavily to shore up their resources: hiring additional developers, training existing staff, and paying for costly software tools and development environments.

The initial build is just the beginning. What follows are the ongoing costs of refining and updating the solution as new technologies and cybersecurity threats emerge. Off-the-shelf solutions come ready to go with the latest features and updates. By choosing an existing product, organisations can leverage the provider’s team of experts who live and breathe the solution.

Maintenance and security: The overlooked giants

Maintaining a homegrown solution is like trying to hit a moving target. It’s not just about keeping things running smoothly, it’s about adapting to a constantly-changing landscape of security threats and technological advancements. Companies often underestimate the sheer volume of work involved in keeping their solution up-to-date and secure.

And it’s not just the work; it’s the specialised knowledge needed to understand and implement the latest security measures. Established and trusted solution providers will have this understanding already.

For example, digital marketing platform Eagle Eye AIR invests 5% of its revenues in improving its platform’s cybersecurity every year. The investment delivers a return on security up to five times higher than the same investment made by an enterprise retailer running a single programme on its own homegrown solution.

Opportunity costs and time to market

The concept of opportunity cost is another conversation that comes up regularly among teams contemplating building their own tools. Building a solution in-house doesn’t just tie up financial resources – it also ties up the time and energy of skilled staff.

Every hour that a team spends on development is an hour not spent on strategic initiatives that could be driving a business forward.

Extended development cycles can be a big trap that cause organisations to miss crucial market opportunities. Technology moves quickly, and the market could potentially move on before the build is complete, blunting an organisation’s competitive edge.

With an out-of-the-box product, organisations get a ready-to-deploy solution that gets them to market faster. That speed translates into immediate benefits, whether it’s staying ahead of competitors or responding to customer needs quickly.

Indirect costs: The intangibles that add up

When organisations build in-house, they are not just paying for development and maintenance. They are also absorbing the costs of project management, internal meetings, testing, and troubleshooting.

There’s also the morale and productivity impact on teams. Nothing drains energy faster than long debugging sessions or development roadmaps that never seem to get shorter. It’s easy for teams to get bogged down in the minutiae of development, losing sight of their core competencies and what truly drives value for their customers.

All indirect costs are minimised when choosing a turnkey solution. Robust, supported platforms allow teams to focus on what they do best – serving customers and growing the business. This not only saves money on development and maintenance; it also preserves team energy and creativity for the tasks that matter most.

The value of expert support and continuous innovation

The value of having access to expert support and continuous innovation can’t be overstated. When building a solution in house, it falls on the development team to handle every challenge. If something breaks or if there’s a need for a new feature, developers are on the hook to figure it out. This often leads to burnout and frustration, especially when resources are stretched thin.

With a professional, prebuilt solution, support teams are there to assist whenever help is needed. Regular updates and new features will also mean the best tools are always available.

Predictable costs and budgeting confidence

One of the most appealing aspects of deploying a market-ready product is the predictability of costs. With an in-house solution, costs can spiral out of control quickly. Unforeseen challenges, feature creep, and the need for additional resources can blow budgets wide open.

Starting with a seemingly reasonable development budget, only to arrive in a financial quagmire months down the line as costs blow out, is a precarious position for solution buyers to be in.

Commercial products offer clear and predictable pricing structure. That not only makes budgeting easier but allows organisations to plan finances with confidence, knowing that the investment will yield a strong return without the risk of unexpected costs.

Focus on core business objectives

The most compelling reason to choose an off-the-shelf solution over building an in-house solution is that the former allows teams to focus on core business objectives. Every business has a unique mission, and time and resources are best spent pursuing that mission, not getting mired in building and maintaining complex solutions.

Freeing up teams to focus on what they do best, whether that’s providing exceptional customer service, innovating new products, or expanding into new markets is often far more preferable for organisations than having teams locked up in long development cycles.

The last 10% takes 90% of the time

According to Melanie Mitchell, computer scientist and Professor of Complexity at the Santa Fe Institute argues that “the first 90 percent of a complex technology project takes 10 percent of the time and the last 10 percent takes 90 percent of the time.”

Teams may have exceptional product owners, product managers, engineers, all of whom capable of building a bespoke solution for loyalty, promotions and personalisation. But having them do so may not be the best use of their time.

However, building that sort of technology is the core focus of a dedicated solution provider. Finding an expert in all the aspects of Melanie Mitchell’s “last 10 percent” is a gamechanger.

The smart choice for reducing TCO

The decision to deploy a ready-made solution versus building an in-house solution can often pose challenges for technology decision makers, but the benefits of doing the former are clear.

Considering all the factors – development time, maintenance, security, opportunity cost, indirect costs, and the ability to focus on core business objectives, choosing a packaged loyalty, promotions and personalisation solution is simple. A pre-built solution reduces total cost of ownership for an organisation and positions it for long-term success.

When it comes to loyalty and personalisation suites, many customers make this choice with solutions like Eagle Eye AIR. The peace of mind, reduced costs, and the ability to stay agile and competitive are all powerful advantages that offer benefits.

It’s not just about saving money; it’s about making a strategic investment in a business’s future.

The post Build or buy your next tech solution? appeared first on TechWire Asia.