Founding Engineer

Description

Every investment firm assembles its own partial view of the world from government data, company filings, private data, news, research, websites, PDFs, spreadsheets, and conversations. Keeping that view current is constant work. Teams must find where information is published, determine what it means, reconstruct the historical record, and adapt whenever a source, format, or definition changes—all while knowing that every other serious investment firm is repeating much of the same work, and possibly doing it faster and more comprehensively, costing them alpha.

The company uses AI to turn these disparate sources into a continuously updated, source-backed model of an industry. It monitors them for new data and research, then reconciles each update with the existing record. Every value traces back to its source, and new information flows through the model — and out to customers — in real time.

Today, the company covers healthcare. It will expand sector by sector until the model spans the global economy.

Who we are

Adam Ron, cofounder and CEO, spent seven years in equity research at Bank of America including leading the sector in healthcare the company covers today. Before that, he started his first company (GoPuff competitor) in college out of his dorm room.

Cameron, cofounder and CTO of the company. I was a founding engineer at Petal and Pinwheel, joining both as the first founding engineer, so I know what its like to join early. At Petal, I built a cash-flow underwriting engine used to evaluate more than 100,000 applicants. At Pinwheel, I built infrastructure that kept income and employment data current for hundreds of thousands of people across hundreds of payroll providers.

We started the company with firsthand experience on both sides of the problem: doing the research and building financial systems from scratch.

Revenue tripled over the last three months and is on track to triple again this quarter. Teams at 5 of the top 10 investment banks and 8 of the 20 largest hedge funds already use the company.

One of the few ACTUALLY AI native companies

We believe agents can do far more than most companies trust them to do. At the company, our backend IS an agent (OpenClaw/Hermes/Eve) given enough tools and skills to run an equity research firm.

Most agent systems place an LLM inside a workflow whose steps are still hard-coded, or use very narrowly defined agents to handle particular tasks. We take the opposite approach. We give our agent system control of our entire research process end to end, letting it determine the steps required, and use deterministic code as tools.

We believe the answer is often less workflow code, not more. When an information source changes or a process stops working, we can update a prompt, skill, or tool instead of adding another branch to the workflow, or the agent can just figure it out.

Today, the company’s agents handle thousands of jobs each day this way. This does not make engineering less important. It changes what we need to engineer. Making agents dependable requires reliable tools, durable execution, strong observability, and clear controls around what they can do. It also dramatically increases the scope of what we can handle.

Responsibilities

• You’ll work across the entire stack. Your work will include:
• Shipping frontend and customer-facing features, including real-time systems that update customers as information changes.
• Creating data pipelines, tools, and infrastructure that let our internal analysts add new sources and maintain sector coverage faster and more reliably.
• Improving our tools and optimizing skills that make Hermes agents more capable and reliable.
• Owning work end to end: inspecting the data, using and QAing what you build, instrumenting it, shipping it, and fixing it when it breaks in production.
• What your first 30 days will look like
• On your first day: Add a new information source to the company.
• In your first week: Ship an improvement to the speed or reliability of the company’s agentic parsing and extraction systems.
• Within your first 30 days: Build the company’s company ontology and ship the first version of Earnings Center on top of it.
• Who we’re looking for
• We do not expect you to have already done every part of this job. We want someone with real depth in backend engineering, data infrastructure, applied AI, or product engineering—and the openness to work across the rest, including unfamiliar infrastructure and scaling problems.
• You should have:
• Built, shipped, and operated something real in production.
• Used agents to accomplish real work, and are excited about orchestrating coding agents in parallel
• Good judgment about when an agent should reason and when deterministic software should take the wheel
• Deep agency and ownership. You want responsibility for the whole result, not just your part of the implementation.
• Bonus: Capital Markets Experience
• Our stack
• Openclaw/Hermes/Eve grade agent harness, Python, FastAPI, DBOS, TypeScript, React, Postgres, Neon, MotherDuck, Turbopuffer, dbt, AWS, and Cloudflare.
• Experience: Any (new grads ok)
• Visa: US citizen/visa only