Technologists Krish Chelikavada and Keon Woo Kim had to make a tough call that many startup founders hope to avoid: they walked away from a working product and shifted their entire company in a new direction.
Their first startup, 0xPass, aimed to improve wallet infrastructure for the crypto space. The team saw initial traction thanks to a strong technical foundation and early funding, but as the crypto market began to lose momentum, they realized it was time to shift their approach.
Krish and Keon turned their attention to AI and relaunched as Om Labs, a venture-backed startup operating at the intersection of AI and infrastructure, with a focus on tools that integrate into existing workflows and automate repetitive work. Their first products, Support3 and Jina, are AI agents designed to handle technical support and software testing with minimal human effort, helping teams move faster and freeing engineers to focus on higher-level tasks.
Read on for a closer look at how Krish Chelikavada and Keon Woo Kim leveraged their crypto beginnings to build practical AI tools with Om Labs and redefine what product-market fit looks like in the generative AI era.
Balancing a Builder's Mindset with Deep Technical Skill
Om Labs is shaped by the complementary strengths of its co-founders, who both studied computer science at NYU but followed different professional paths before eventually reuniting.
Krish developed an entrepreneurial mindset early on, spending his free time at college on fun personal projects like building Android apps and a tutoring marketplace for students. He went on to work as a software engineer at Oracle before moving to product management at Microsoft. After earning a master's in management science and engineering at Stanford, he focused on building full-time.
Keon, on the other hand, focused on technical depth. He began his professional career at Uber, where he single-handedly led the rebuild of the company's driver incentives system, a core project affecting gig workers in California and New York. Outside of his engineering work, he made significant contributions to open-source reinforcement learning projects, with his codebases amassing a combined total of over 48,000 stars on GitHub. He has also authored several books on deep learning and reinforcement learning, which are now used across university courses in Korea.
Together, they bring a unique combination of product sense and engineering expertise, which proved to be especially valuable as they entered the complex, fast-moving worlds of AI and crypto.
Launching a Crypto Product in a Shrinking Market
When Krish and Keon first teamed up, they set their sights on the crypto space with the aim of building tools that could support its growing infrastructure.
In late 2022, they began developing 0xPass, a system designed to simplify how applications generate, store, and use cryptographic keys. It was built to reduce the operational burden for teams managing large volumes of wallets by distributing key shares across multiple network nodes, thereby avoiding a single point of failure, improving security, and lowering backend complexity.
Early performance benchmarks showed strong results, with the platform achieving sub-second response times and outperforming widely used tools like NEAR. But by the time they were ready to launch, the crypto market entered a sharp downturn, with multiple major assets like Bitcoin seeing significant price drops and tightening regulations making it harder for the technology to gain real-world adoption.
As a result, investor interest in crypto projects declined considerably.
Despite these challenges, Krish and Keon moved forward with the project. They developed a strategy that involved pitching 0xPass not only to crypto-native users but also individual investors, framing the product as part of a larger vision for secure infrastructure. This approach helped them raise $1.9 million in pre-seed funding and attract clients managing over 10 million wallet connections during the first year.
Still, by late 2024, the crypto market continued to struggle. Despite having a foothold in the space, they ultimately made the decision to pivot and shift their focus to a field that was growing quickly and better aligned with how they wanted to build: AI.
Relaunching as Om Labs: Building AI Tools That Streamline Manual Work
By the time Krish and Keon turned their attention to AI, more and more companies were adopting the technology to automate their internal operations, with reports estimating that 34% of business tasks were being automated in some form.
Their first product in the AI space was Support3, an automated technical support agent for early-stage and developer-focused companies that communicate with users through platforms like Discord, Telegram, and Slack. Support3 automatically monitors conversations in these channels, responds to technical queries, logs relevant information across all threads, and summarizes ongoing discussions. By surfacing real-time user feedback and informing technical teams of recurring issues, Support3 allows companies to stay on top of support and improve their customer experience without having to track every conversation manually.
They followed this up with Jina, a tool that automates quality assurance testing by simulating how a real user would interact with an application to help engineering teams catch problems earlier and release more confidently. Instead of writing test scripts, users simply describe the desired user flow in plain language, and Jina quickly navigates the interface, identifies bugs, and generates detailed reports on its findings.
Om Labs is building tools that eliminate common development bottlenecks, speed up release cycles, and improve accuracy across key workflows.
Building the Next Generation of AI Employees
Their long-term goal is to build what they call "AI employees" that can manage complete workflows without manual oversight and work as integrated teammates rather than just add-ons. Their goal isn't to replace entire teams but to offload the manual tasks that slow teams down, allowing engineers to focus their energy on more sophisticated work like architecture, feature design, and long-term product planning.
Krish Chelikavada and Keon Woo Kim are using Om Labs to offer a practical glimpse into their vision of ambient AI: always on, always learning, and deeply embedded into how teams work.
ⓒ 2025 TECHTIMES.com All rights reserved. Do not reproduce without permission.
