you.com is an AI search infrastructure company that enables AI agents to access real-time web search through APIs. Pulse 2.0 spoke to You.com CTO Saahil Jain to learn more.
Sahil Jain Biography
Could you tell us more about your background? Jane said:
“I completed my undergraduate studies in computer science at Columbia University, then went on to my master’s program at Stanford, where I was lucky enough to join Professor Andrew Ng’s machine learning group. My research focused on deep learning and NLP in resource-constrained environments, particularly healthcare. The main challenge is that most ML research assumes an abundance of labeled data, which is often absent in hospital and clinical settings. It has been presented at ML conferences and has accumulated over 10,000 citations, which I am proud of. I am also grateful to be part of Stanford’s open source movement, which is working to democratize AI research in healthcare.
I joined You.com in December 2020 as a founding engineer. CEO Richard Socher was convinced that language models were going to fundamentally change search. I also agreed with that. He spent the next five years building a search and AI infrastructure from scratch, rising to become head of AI and was recently named CTO. ”
Establishment of a company
How did the idea for the company come together? Jainism said:
“Back in 2020, Richard realized that language models were becoming more powerful and focusing on search as an ideal target for AI disruption (something most people disagreed with). I happened to agree and joined You.com in December 2020 as a founding member of the engineering team.”
“We started deploying language models at scale in the search context as early as September 2021, when everyone was still used to getting a list of 10 blue links. We were completely rethinking search and using AI to provide summarized and synthesized answers rather than just links.”
“The general availability of ChatGPT in November 2022 changed everything. Everyone was suddenly excited about conversational AI. Everyone quickly realized the fundamental problem of knowledge blackout. Because it was trained on old data and completely disconnected from the web, it couldn’t know what the weather was like today or the score of last night’s game.”
“We set out to solve that problem, and just weeks after launching ChatGPT, we became the first company to connect LLM to real-time web data using verifiable citations. This solved two critical problems at once: outdated information and AI illusions.”
Main products
What are the company’s core products and features? Jain explained:
“We’ve built a search infrastructure that makes AI agents accurate and reliable. Our main product is a suite of APIs for developers. We provide search, content, and research APIs that enable teams to build agents that don’t hallucinate, don’t get stale, and don’t break under production workloads. Our search APIs deliver up to 10 searches per month for companies like DuckDuckGo, Windsurf, Harvey, and Databricks. Processing over 100 million queries. In our benchmarks, we show superior performance against both Google’s Search API and the Bing (now deprecated) API when it comes to accuracy and speed. ”
Evolution of the company’s technology
How has the company’s technology evolved since its launch? Jain pointed out:
“While we started as a consumer AI search engine, we realized there was a bigger opportunity in building a search infrastructure that powers AI agents in a trusted web context.”
important milestones
What is the company’s most important milestone? Jain says:
“We have made great progress over the years, but there are a few moments that I am particularly proud of.”
“In 2022, we first integrated LLM with real-time web search, solving the knowledge blockage problem that makes chatbots unreliable. Over the next few years, we built an API infrastructure that now serves over 1 billion queries per month.”
“In February 2025, we launched ARI, a deep research agent that leverages search APIs to synthesize over 500 sources into comprehensive reports with verifiable citations, allowing knowledge workers to use them for real-world tasks. And in September 2025, we raised $100 million in Series C and became a unicorn at a $1.5 billion valuation.”
I’m particularly excited about the Research API, which we launched in February of this year to expand our deep research capabilities. It ranked #1 on DeepSearchQA, a 900-prompt benchmark that evaluates difficult multi-step information-seeking tasks across 17 disciplines. We’re proud to announce that our Research API achieved an accuracy score of 83.67%. This is the highest value ever reported in the industry. ”
Customer success stories
Can you share any specific customer success stories?Jain highlighted,
“A partnership worth mentioning is Databricks, who have integrated our search API directly into the Unity catalog. This means that companies building AI agents on Databricks can now access real-time web data via their API without leaving the platform, as their credentials are protected, requests are audited, and access is managed through Databricks’ built-in controls.”
“It would be remiss of us not to point out that You.com is one of the default search providers for OpenAI’s openweight GPT model. DuckDuckGo and Windsurf are also major adopters of our Search API. They use the Search API to provide visitors with the latest news and coding agents with access to current repositories and documentation, respectively.”
Total addressable market (TAM)
What total addressable market (TAM) size is the company pursuing? Mr. Jain assessed:
“We are building infrastructure for the agent era, where AI agents will browse the internet more often than humans. Current search infrastructure is built for humans to click links. AI agents need something completely different: deep contextual information from both private and public data to make decisions and take action.”
“TAM is essentially ‘any company that builds AI applications that require accurate, up-to-date information.’” That ranges from everything from legal research (Harvey) to developer tools (Windsurf) to enterprise analytics (Databricks) to search engines (DuckDuckGo). As AI becomes critical infrastructure rather than new, the need for accuracy and reliability will become non-negotiable. ”
future company goals
What are the company’s future goals? Jain concluded:
“Just as AWS has become the default infrastructure for cloud computing, we are focused on becoming the default search infrastructure layer for enterprise AI. In the short term, that means expanding our API services as more companies realize they need infrastructure they can trust rather than trying to build their own.”
“In the long term, it is important to establish new standards for accuracy and reliability in AI systems. As agents handle increasingly important decisions, their underlying infrastructure must be flawless. We want ‘powered by You.com’ to be the hallmark of the highest reliability.”
