AI Engineer

Who are we? Gain is the Private Markets Super App, a connected global platform transforming how leading investors, advisors, and corporates source, evaluate, and execute deals. By combining investment-grade intelligence, AI-powered workflows, and proprietary data in one unified platform, Gain enables top deal teams to move faster, reduce risk, and build conviction with confidence. Gain is trusted across more than $1 trillion of private capital, used by 100% of MBB and Big Four firms, and by 80% of the world’s top 20 M&A advisory houses. Recognized for fast innovation and impact, Gain has been named both US and EU Data Provider of the Year by PE Wire, awarded Global Financial Market Review’s Best Use of AI in Finance, and ranked in Sifted’s Top 100 Fastest-Growing Companies, the Forbes 1000, and the Deloitte Technology Fast 50. Gain operates globally with offices in New York, London, Amsterdam, Frankfurt, Warsaw, and Bangalore. How we work • Predicting what isn't published. Many private companies publish little or no financial detail. Across a database of more than 5 million companies, we predict revenue and EBITDA from sparse labels, and we predict whether and when a company will be bought or sold, which is a time-to-event problem with censored data. • Agents on a messy web. Our agents search, navigate and extract company data points from sources that contradict each other, then decide which source to trust before committing to an answer. • Quality that holds while everything underneath moves. We're provider-agnostic. Model choices follow the evals, and we switch LLMs when the numbers say so. That only works if the evals are trusted. What will you do? The work is roughly half LLM agents and half custom ML, leaning slightly toward agents. • The design of our agent pipelines: architecture, tools, and which model for which task. You'll shape the technical direction with our AI Lead. • The evaluation behind agents and models: offline reference sets, live production evals, and the call on whether an improvement is real. • Our prediction models for financials and deals: features, validation, retraining, and knowing when a simpler model wins. • Getting work into production: trace where pipelines fail and fix the weakest links first, with compute, memory and model cost in view. • Raising the bar in the team through design reviews, pairing, and an eval-first habit that sticks. In your first six months you'll own the evals for our agent pipelines, propose where we invest next, and ship measurable improvements to both the agents and the prediction models. Who are you? • You've built and owned ML systems in production for years, typically 6+ in ML or software engineering. • In the last 1–2 years you've shipped LLM agents or pipelines that real users depend on, not demos. • You have modelling judgement. You've dealt with leakage, distribution shift and calibration on real data. • You measure before you claim. You can build an evaluation set and tell an honest improvement from a lucky one. • You care about the user outcome more than the technique. Ideas without a clear customer benefit don't get built here, however interesting. • Nice to have: time-to-event modelling, noisy or conflicting data sources, financial or private-company data. No finance background needed. The team You'll join our Data & AI team of six and report to Renan, our AI Lead who's been building production models for >10 years. Our engineering team runs the infrastructure, and you shape what runs on it. The stack • Core: Python (async, fully typed with Pydantic) and FastAPI • LLMs: multiple providers behind our own provider-agnostic layer, with Pydantic AI for typed agents. • Evals and observability: Pydantic Evals and Logfire • Data: PostgreSQL with p