Taken from Tech in Asia
In partnership with 
The journey of a tech startup, especially in Southeast Asia, is rarely a straight line to success.
One of the first hurdles startups face is proving their concept’s viability. Convincing potential clients, particularly large enterprises accustomed to established solutions, to take a chance on emerging, untested technology can be daunting.
“A lot of enterprises want to try products before they deploy it because once it’s deployed, it means you’ve already set everything, so change is very difficult,” explains Jennifer Zhang, CEO of Singapore-based AI startup Wiz.ai.
This need for validation translates into a demand for proof-of-concept (PoC) projects, where startups demonstrate the functionality of their product and its business value in a real-world setting.
This is especially crucial for Wiz.ai, whose focus on conversational AI – specifically AI agent solutions for customer engagement – introduces a level of novelty that is both intriguing and intimidating for potential clients.
The PoC serves as a bridge between startups and clients, allowing the former to demonstrate the real-world applications of their solutions to the latter.
“The purpose of a proof-of-concept is to test several things, but the primary goal is to showcase the technology’s feasibility or advantage,” explains Zhang. “We always ask ourselves, ‘If we deploy this solution on a small scale, how can we ensure it can effectively handle a much larger, real-world environment when fully implemented?’”
Wiz.ai’s technology aims to make customer service operations more efficient with the use of AI-powered virtual agents. In order to prove the viability of its solution, the firm not only has to showcase the accuracy of its AI but also demonstrate how its technology can integrate with existing systems and workflows.
Jennifer Zhang, CEO and co-founder of Wiz.ai
Designing and executing a successful PoC involves many considerations and complications, ranging from determining the actual scope of work needed to aligning with technical requirements. It aims to ensure that a solution can meet the demands of commercial implementation.
“We use the PoC phase as a learning ground to understand our customer needs and adapt accordingly,” adds Zhang.
For Wiz.ai, this has involved demonstrating to a bank how its AI model can accurately recognize and understand various Southeast Asian languages and showing a hospital how its AI-powered call center operations could automate appointment scheduling, improving operational efficiency and cost savings.
“You need to monitor everything, make changes to the workflow as necessary, to understand and build that trust with the customer, and also learn more about their processes and infrastructure,” Zhang says.
While securing enterprise clients can be challenging, working with government agencies adds a level of complexity. The stakes are often higher, given that public data and sensitive information may be involved, and the procurement process is much more stringent.
Unlike private enterprises, which are usually open to testing and iterating on a new solution, government agencies often require a fully developed, ready-to-deploy product, even for a PoC. They also have stricter requirements around security, data management, and integrations, which can be overwhelming for a young company.
“You can’t experiment or collaborate on something that’s still in its early stages. The expectation is a mature, functional product that’s ready to be implemented,” Zhang notes.
This posed a challenge for Wiz.ai when it was onboarding government partners, as like most startups, it thrives on a model of iterative development and close collaboration with clients. However, the firm was able to work with the Tech Acceleration Lab (TAL) to overcome this hurdle.
A part of the Singapore Infocomm Media Development Authority’s Accreditation program, TAL aims to accelerate the PoC process to help startups work with government clients. In turn, it provides government organizations a more effective way to access new and innovative solutions.
“IMDA works with you to make sure that everything’s good to go, you understand how the government system is, what their security requirements are, and what the integrations look like,” Zhang adds.
With TAL’s guidance, Wiz.ai was able to secure its first government project with Singapore’s Ministry of Social and Family Development (MSF). The ministry had multiple hotlines available, and citizens were sometimes unsure of which number to call. Wiz.ai’s Talkbot solution addressed this issue by redirecting calls to the right places.
This led to a 20% reduction in the number of calls to MSF’s general enquiry hotline, and it enabled Wiz.ai to gain a deeper understanding of the government’s specific needs and requirements. It has also opened doors for the company to develop more PoCs with other government agencies, culminating in a “whole-of-government” project that spans multiple entities.
You can’t experiment or collaborate on something that’s still in its early stages. The expectation is a mature, functional product that’s ready to be implemented.
“IMDA sets a good standard for many enterprises and organizations in the ecosystem,” Zhang says. “For example, in telecoms and banking where standards are stringent, our collaboration with IMDA serves as a marker of quality of our service and product.”
TAL also supports startups in building PoCs for enterprise customers, something that Wiz.ai has benefited from as well.
“Working with TAL also proves the standard of data security provided in our solution and, in turn, results in confidence from enterprises all over the region,” Zhang adds.
While the immediate benefits of a successful PoC and deployment are clear – increased credibility, new clients, and expanded market reach – the impact extends beyond the individual company.
The rigor involved in working with clients in regulated sectors like government sets a higher bar for the AI industry as a whole. Companies will need to prioritize security, data privacy, and robust integration, which leads to more mature and reliable solutions.
The Wiz.ai team
Furthermore, successfully navigating the complexities of government procurement can serve as a blueprint for other companies, fostering an ecosystem where startups and government agencies work together to develop and deploy cutting-edge solutions. To date, TAL has supported 102 startups and facilitated 130 PoCs, creating new opportunities for collaboration. It has also reduced the time taken for PoC cycles by 4x to 4.5x, bringing it down to just one to two months,
“After successfully providing solutions to government agencies, a startup can confidently say they’ve met the highest standards of security, data management, and solution design required,” Zhang says.
She believes that securing a project with the government can boost a company’s credibility and unlock opportunities in new markets as well as with larger enterprise clients. IMDA has found that 40% of the PoCs under TAL’s enterprise scheme and 50% of those under its government arm have been converted to full projects.
“If you run the whole [PoC and deployment] process well, it has a wider impact on the entire industry,” she adds. “It gives confidence for these startups and agencies to go a lot further.”
Wiz.ai is part of the IMDA Accreditation program, which was launched in 2014 to accelerate the growth of promising Singapore-based enterprise tech firms. The program also includes the Tech Acceleration Lab, a bespoke initiative where portfolio companies are guided through the PoC and deployment process end to end.
Find out how you can accelerate your PoC-to-production deployment on IMDA’s website or email the TAL team now at tech_acceleration_lab@imda.gov.sg.
Copyright © 2023 WIZ Holdings PTE. LTD. All Rights Reserved
