More companies are jumping on the artificial intelligence (AI) bandwagon, but many are still figuring out how to connect technology to business outcomes. For Singapore-based innovation company AngelHack, hackathons are the key to solving problems.
In an interview with Computer Weekly, Ryan Chew, director of AngelHack DevLabs, said that despite the rapid growth in AI adoption, enterprises continue to struggle with using large-scale language models (LLMs) to build AI applications that fit their business contexts and workflows.
At AngelHack, which runs a variety of developer and innovation programs, our expertise in hackathon planning is in high demand. Rather than relying on external consultants who don’t understand the company’s operations, AngelHack runs in-house hackathons to teach non-technical employees, such as marketing and finance teams, how to build their own AI-powered applications.
“Many companies like to hire consultants from outside parties who say, ‘Try AI,’ but they don’t know the workflow, they don’t have privileged information, and they don’t have a presence in their day-to-day operations,” Chu explained. “All you have to do is spend a week teaching the people who are the actual stakeholders. They know their workflow best, so they can produce better results.”
Boasting a global database of developers, AngelHack also hosts large-scale external hackathons for tech giants like Microsoft to drive adoption of tools like Copilot.
These events are used not only for product adoption, but also for recruitment efforts and employer branding. Local telco Singtel recently reached out to AngelHack to showcase its investments in data centers and AI capabilities, while government initiatives such as the Defense Science and Technology Agency’s BrainHack use these competitions to identify and hire top technology talent.
If your hackathon results in a commercially viable project, AngelHack can also help bring your project to market. In Singapore, the company is working with Digital Industry Singapore (DISG) and AI chip giant Nvidia to run an accelerator program to turn prototypes into scalable businesses.
And for companies that lack the in-house engineering capacity to bring their ideas to life, AngelHack’s DevLabs division steps in to build minimum viable products (MVPs) and custom software applications to help companies ease their transition to cutting-edge technologies like AI and blockchain.
organizational challenges
As AI gains the ability to write its own code, the role of software engineering is also undergoing a major transformation. Chew calls this the “surface area” problem. How should companies organize their engineering departments when a single developer using AI can now cover the amount of work that previously required a team of five?
“Do we have enough capabilities for everyone to work on? If so, we need to redesign the organizational structure within the engineering team to clarify roles such as audit and PR reviews.” [pull requests]”.
While some big tech companies have used the efficiency of AI to justify mass layoffs, Chu sees technology as a multiplier that should be used to increase a company’s output rather than reduce its workforce. “If we have AI that empowers engineers, we should be able to produce more without cutting people,” he said.
Additionally, there are governance, risk, and compliance issues that companies must address. “There is no AI that is 100% correct, and no human being is 100% correct. I can hold humans responsible, but I cannot hold AI responsible,” Chu pointed out. As a result, much of the enterprise AI development now facilitated by AngelHack is focused on internal workflows rather than risky external-facing applications.
AngelHack is currently riding a wave of global growth. The company, which has about 70 employees, is seeing increased demand across its academies, hackathons and custom solutions business. In particular, Chu said more small and medium-sized enterprises (SMEs) are working with DevLab to understand how AI can improve internal efficiency or integrate into existing products.
The company is also expanding its footprint by organizing hackathons for customers in Australia and Brazil. The company also plans to hold hackathons in multiple cities across the United States later this year.
As the barriers to technology development continue to fall, Chu noted that companies like AngelHack’s appetite for expert facilitation shows no signs of slowing down. “Even if you know how to cook, sometimes you want to eat at a restaurant when cooking is so easy,” he quipped.
