Asyn is building proprietary AI infrastructure for investment firms that manage the $161tn global bond market. A fixed income firm's coverage, margin and AUM have only ever scaled with headcount. Asyn replaces that constraint with purpose-built agents that reduce the time to decision in the investment process by delivering expert human-grade work. Founded by a fixed income expert with 13 years managing $20bn+ portfolios and a second-time VC-backed technical founder, we are live with a handful of regulated investment managers and growing fast.
The role
You'll be one of our first technical hires, working directly with the founders to build world-class agentic decision infrastructure for fixed income. You'll be in the driver's seat, shaping the product from day 1: the agentic research/decision systems at the core, the evaluation infrastructure that proves they work, and the continual learning loop that makes them compound over time. This is a role for an elite AI engineer who wants depth, breadth, ownership, and a direct mandate to build frontier agentic systems in high-stakes finance.
What you'll do
- Design and build agentic harnesses for deep-research workflows: multi-step analyses over deals, issuers and markets with tool use, memory and orchestration.
- Build our eval stack: custom benchmarks for financial reasoning tasks, model/agent evaluation pipelines and regression tooling that let us ship confidently.
- Post-train and fine-tune open-weight models on proprietary workflows and data, so each customer's edge compounds inside its own walls.
- Work alongside fixed income analysts, traders and portfolio managers, learning about this fascinating field and co-designing solutions for them.
- Work with financial datasets - bond pricing, issuance calendars, market data, building the analytics and data pipelines the agents depend on.
- Shape engineering culture, tooling and architecture from day one.
Some skills that matter
- Experience building agentic harnesses or applications.
- Prior experience setting up model and/or agent evals and building custom benchmarks.
- Hands-on post-training and/or fine-tuning of open-weight models.
- Solid foundations in classical ML and deep learning.
- Solid fundamentals in statistics, data analytics and data science concepts.
- Strong software engineering fundamentals - you write production-quality code and can own systems end-to-end.
- A fearless, can-do mindset coupled with enough humility to know your limitations and work towards overcoming them.
Nice to have
- Domain experience or genuine interest in finance - especially trading, portfolio management or fixed income.
- Original research published at leading AI conferences.
- Former (technical) founders or early engineers in startups will be regarded highly.
What we offer
- A frontier problem: proprietary AI for one of the largest and least-automated markets in the world.
- Direct exposure to customers, investors and technical / product decisions.
- Competitive salary based on expertise and early-stage stock options.
- Office in central London with a bunch of perks.
#J-18808-Ljbffr
