The enterprise software model was built on siloed apps and integrations, and is beginning to show its age. According to a recent report, the average company now uses more than 100 SaaS applications, each with its own workflows, interfaces, and maintenance needs.
Krish Chelikavada and Keon Woo Kim are two technologists building an alternative through Om Labs, a startup creating “AI employees” that function as autonomous agents and plug into internal systems to carry out real work on their own. From customer support to product testing, these AI agents can interact with users, troubleshoot, and execute tasks end-to-end.
One key example of this framework is one of their latest platforms, Jina, which simulates end-users to test software automatically with minimal friction before shipping. This tool goes beyond streamlining workflows and stands in for parts of the traditional software stack, allowing teams to shift from operating applications to managing automated, intelligent systems.
Chelikavada and Kim specialize in building real, deployable AI agents that act more like coworkers than chatbots. Here’s a closer look.
Two builders with a track record of shipping real systems
Chelikavada holds a bachelor’s in computer science from NYU and a Master of Science in management science and engineering from Stanford. He worked at Oracle before launching Om Labs, but it was his knack for experimentation that put him on his current path.
“I’ve always been a builder and tinkerer of apps and projects,” he explains. “In college, there was a time when I built Android apps almost daily. I just felt fulfilled by it.” His instincts helped to shape Om Labs and its first products, from initial design to fundraising.
His co-founder, Kim, has deep systems engineering experience. At Uber, he built backend pipelines that handled more than $200 million in monthly driver payouts. He has also authored two technical books on deep learning and reinforcement learning which have been adopted by universities in Korea.
Kim’s mix of backend expertise and academic credibility complements Chelikavada’s product-first approach, a combination proven in practice.
Chelikavada and Kim built 0xPass in late 2022 to help manage the digital “keys” people use to access and secure cryptocurrency wallets. What set 0xPass apart was its ability to verify and manage these keys in less than a second, much faster than other solutions.
Shortly after launch, Chelikavada and Kim raised $1.9 million in pre-seed funding.
From crypto success to a new vision in AI
After building 0xPass into a fast and reliable key management system, Chelikavada and Kim made a deliberate decision to pivot Om Labs toward artificial intelligence. Despite having a foothold in the crypto infrastructure market, they saw a larger opportunity emerging.
“We turned down potential seed round discussions to trust our conviction that it was time to build in AI,” Chelikavada recalls. “Despite having the best key management solution in crypto, we made a call to move to AI, starting with a support product. We looked at it as a massive opportunity.”
This shift aligned with a broader trend. The crypto market was beginning to stall while generative AI adoption was rising rapidly, and the traditional SaaS model began struggling to match the agility and reduced overhead of AI-driven solutions.
Rather than adding more apps, the founders envisioned AI systems that could operate autonomously within existing workflows.
Jina: an AI employee with real applications in real workflows
One of the highlight projects that Chelikavada and Kim developed under Om Labs was Jina, an AI-powered synthetic user that autonomously tests software.
When given a task description, Jina can click through interfaces, simulate real user behavior, and report bugs and usability problems before products reach customers. It produces every report in natural language and with plenty of screenshots, making it easy for non-developers to spot potential issues.
Jina runs continuous tests and automatically adapts to new app versions in a way that a human quality assurance team would, helping development teams catch regressions and reduce manual QA efforts so they can deliver better software faster than ever before.
This framework illustrates Om Labs’ broader mission of embedding AI employees into live systems and performing high-cognition work. For Chelikavada and Kim, Jina is just the beginning.
Building toward an AI workforce
Krish Chelikavada and Keon Kim envision a future where AI employees can go beyond specialized tools to operate continuously within enterprise systems, without waiting for prompts.
This approach, which Kim calls “ambient AI,” frames agents as infrastructure. They become a source of truth and action that works in the background to improve outcomes and reduce overheads.
As AI moves from assisting humans to acting autonomously, the implications for enterprise software are becoming clearer. Om Labs is developing full-stack agents designed to integrate with live workflows and play a key role in reshaping how businesses manage operations and interact with users.
It’s an early signal of a broader shift towards a future where AI employees take on real responsibilities while human teams are freed up to focus on more creative and strategic work.
Disclaimer:
Investing involves risk and your investment may lose value. Past performance gives no indication of future results. These statements do not constitute and cannot replace investment advice.
VentureBeat newsroom and editorial staff were not involved in the creation of this content.
