# Exa hires Jeff Pinner after his Robinhood CTO role > Source: > Published: 2026-08-23 00:12:02+00:00 # Exa hires Jeff Pinner after his Robinhood CTO role **Pinner joins after Robinhood separated him from its CTO role in May, as Exa scales the search infrastructure behind AI agents.** By [RuntimeWire Staff](/author/runtimewire-staff) ยท Published Primary source: [X](https://x.com/WilliamBryk/status/2090879068480938003) ## Why it matters Bryk is pairing Exa's founder-led search research with an operator experienced in high-volume platforms, a necessary step after raising $250M to make agent search core infrastructure. Exa co-founder Will Bryk announced Jeff Pinner's appointment as chief technology officer, putting an engineer who previously ran technology at Robinhood and Lyft in charge of scaling the web-search infrastructure Bryk and co-founder Jeff Wang started building five years ago. Bryk announced the appointment on August 22. Pinner is arriving as [Exa](https://exa.ai/?ref=runtimewire) tries to turn its founders' technical thesis - that AI agents need a search engine built around their workloads - into infrastructure capable of serving a much larger market. That shift requires a different kind of engineering organization. Exa is training retrieval models, operating its own web index, running a GPU cluster and serving search under tight latency requirements. Exa also wants to support longer, more computationally expensive research jobs alongside fast API calls. Pinner's experience operating consumer platforms at scale fits the work Bryk has laid out for Exa's next phase. Pinner's move follows a short tenure as Robinhood's CTO. [Robinhood hired him as its first CTO](https://robinhood.com/us/en/newsroom/robinhood-welcomes-jeff-pinner-as-chief-technology-officer/?ref=runtimewire) in August 2024, after he had worked as a distinguished engineer in Cruise's AI and robotics organization. A [regulatory filing](https://investors.robinhood.com/static-files/87269e46-3e7c-411d-a876-de0ad4b0a7b4?ref=runtimewire) shows that Robinhood separated Pinner from the CTO position on May 7, 2026, with benefits applicable to a termination without cause. Before Cruise and Robinhood, Pinner spent almost a decade at Lyft. He worked on Lyft's engineering infrastructure and rideshare marketplace before serving as CTO. He holds a bachelor's degree in mechanical and aerospace engineering from Princeton and a master's degree in aeronautics and astronautics from Stanford, according to Robinhood's public biography. ### From a Harvard experiment to an infrastructure bet Bryk and Wang founded Exa in 2021 under the name Metaphor. The pair met at Harvard, where Bryk studied computer science and physics, researched human-AI interaction and led the undergraduate robotics club. He later became an early engineer at enterprise AI developer Cresta. Wang studied computer science and philosophy before spending three years building data and web infrastructure at Plaid. Their initial idea grew from Bryk's frustration with conventional keyword search and the capabilities that appeared with GPT-3. Bryk and Wang believed embeddings could retrieve webpages by meaning rather than by matching a user's exact words. They launched an early search engine in 2022 and [renamed Metaphor to Exa](https://exa.ai/blog/announcing-exa/?ref=runtimewire) in January 2024. The timing of Pinner's hire reflects how far that idea has moved from a research project. Exa now sells search, crawling, webpage extraction and research APIs to developers building coding assistants, sales tools, chatbots and other agent products. Exa says its infrastructure powers search for customers including Cursor, Cognition, HubSpot, OpenRouter and Monday.com. Exa claims its crawlers track [more than 500 billion URLs](https://exa.ai/about?ref=runtimewire) and that [more than 400,000 developers use its products](https://exa.ai/blog/announcing-series-c?ref=runtimewire). Those are self-reported figures. They nevertheless describe the operating challenge Pinner has been hired to manage: indexing a changing web, training specialized retrieval systems and returning useful material quickly enough that search does not become the bottleneck inside an agent. ### Pinner inherits a much larger buildout Exa [raised a $250 million Series C](https://exa.ai/blog/announcing-series-c?ref=runtimewire) led by Andreessen Horowitz on May 20 at a reported $2.2 billion valuation. Bryk said the financing would fund new models and [infrastructure capable of processing hundreds of thousands of searches per second](https://exa.ai/blog/announcing-series-c?ref=runtimewire). Exa said at the time that [more than 5,000 organizations used its services](https://exa.ai/blog/announcing-series-c?ref=runtimewire). The round followed an [$85 million Series B](https://exa.ai/blog/announcing-series-b?ref=runtimewire) led by Benchmark in September 2025. Across its disclosed rounds, Exa has raised at least $357 million. That capital puts Exa in the unusual position of trying to build an independent search index while competing against providers that crawl selected pages, aggregate existing search engines or generate answers on top of retrieved sources. The distinction matters to Bryk's pitch. Exa argues that owning the index and retrieval models lets it tune search for agent-specific jobs, including source discovery for coding systems and structured research across people and businesses. Exa's [current product page lists Instant mode at 323 milliseconds](https://exa.ai/products/search?ref=runtimewire), while Exa Agent supports multi-step research using parallel subagents. In June, Bryk and Wang pushed further into that second category with [Exa Agent](https://exa.ai/blog/exa-agent?ref=runtimewire), an API for multi-step web research. The release moved Exa beyond supplying links and page contents into producing structured research outputs, placing Exa closer to the application layer while preserving the underlying search API business. That expansion gives Pinner two engineering problems at once. He must scale the index and serving infrastructure that made Exa useful to developers, while supporting agent products whose quality depends on model selection, source coverage, orchestration and cost control. Each added layer also puts Exa into closer competition with search APIs such as Tavily, crawling systems such as Firecrawl and research products that combine retrieval with generated answers. Pinner has handled versions of this transition before. Lyft required infrastructure that could match riders and drivers in real time while absorbing rapid growth. Robinhood hired him to oversee centralized engineering functions spanning AI, data science, infrastructure and platforms. Exa is smaller, but its architecture carries many of the same demands: high request volumes, continuously changing data and customers that notice immediately when latency or relevance slips.