Frontier Builders
Frontier Builders are creating companies built natively on OpenAI models, where AI runs real work in the real world. They deploy models that write code, serve customers, identify risk, and drive decisions across domains from go-to-market and customer experience to compliance and biodefense.
The next wave of innovation will not come from a single company shipping features. It will emerge from an open ecosystem of builders using frontier models to push the edge of what’s possible.
Kareem
Building Clay
“We realized that GTM professionals had lots of creative ideas for how to grow a company that they couldn’t implement. And in order to stay ahead, they constantly need to be doing things differently. So, we decided to build a tool that gives them the ability to express their creativity and iterate quickly.”
Turning any growth idea into reality
Kareem approaches go-to-market as a creative discipline. He saw that growth was constrained by systems designed to document the past rather than respond to new ideas.
Clay is built on AI agents that autonomously research markets, analyze signals, and generate outreach strategies in real time. These systems do more than surface insights; they act on them, continuously shortening the distance between idea and execution. The only real limit to what’s possible with Clay is a user’s creativity.
Building with OpenAI
Clay builds on the OpenAI platform to turn raw GTM data into structured signals. OpenAI models are used to classify and extract attributes from company websites, enrich and normalize account data, and interpret sales call transcripts for things like intent and competitor mentions, so Clay can generate actionable insights that power downstream campaign, routing, and sales-prep workflows.
Kathleen & Tess
Building Valthos
“As AI advances, life itself has become programmable. Whether these tools are used to usher in a new era of precision medicine or weaponize biology depends on the steps we take today.”
Systems that defend against biological threats
As biology accelerates, the speed of detection and response determines outcomes. Kathleen and Tess founded Valthos to meet that reality.
Valthos operates AI-driven systems that analyze biological sequences as they emerge, assess risk, and prioritize response paths in hours instead of months, strengthening real-world biological defense.
Building with OpenAI
Valthos uses OpenAI as the reasoning layer that ties together biological models, sequence analysis, and decision pipelines. Outputs from specialized bio models are evaluated, compared, and prioritized through test and verification frameworks, enabling rapid in-silico risk assessment and medical countermeasure triage.
Connor
Building Unify
“We’re rebuilding go-to-market from first principles and giving you a system of action to generate pipeline that is repeatable, observable, and scalable.”
The system of action for pipeline
Connor believes go-to-market is fundamentally a search problem. The data exists, but teams lack a way to surface the best leads and action on them immediately. Unify is a system of action for turning signals and outreach into a consistent pipeline, so that the best products win.
Building with OpenAI
Unify relies on OpenAI’s agentic capabilities and tool-calling to execute millions of automated research jobs across your entire market. Always-on agents detect buying signals, qualify accounts in real time, and personalize outreach with live context. This gives reps clear intent and a real reason to engage so every outreach is relevant and grounded in how their product helps.
Christina
Building Vanta
“Security and compliance should be two sides of the same coin. You shouldn’t just check the box for an audit—you should secure your systems first, and let software prove it.”
Helping businesses earn and prove trust
Christina founded Vanta after seeing how compliance became disconnected from real security.
Vanta runs an agentic trust platform that continuously monitors controls, identifies gaps, and fixes issues in real time, turning trust into an always-on system rather than a periodic snapshot.
Building with OpenAI
Vanta uses OpenAI models to reason and translate data across its security and compliance platform. The models map risks, controls, assets, and evidence into a continuously updated system of record, and assist with interpreting regulations, drafting policy changes, and checking them against new requirements, while Vanta’s own workflows manage validation, approvals, and audit-grade reliability.
Jesse & Ashwin
Building Decagon
“Every customer everywhere deserves a concierge experience, 24/7. If you’re the customer of something, it should feel like you’re the only one.”
The AI Concierge for every customer
Jesse and Ashwin are building a new standard for customer experience: the Concierge standard. It's personalized and proactive, respectful and resolute. Decagon is built for business experts who know their customers best, giving them a powerful platform to create, optimize, and scale AI agents that meet the mark every time.
Building with OpenAI
Decagon uses OpenAI models to power AI agents for customer support and operations that can understand inbound requests, decide on next actions, and execute workflows across tools. With OpenAI’s APIs, these agents can reason through complex situations, follow business rules, and improve over time from real-world interactions, delivering reliable and safe experiences at scale.