Role overview
We're looking for an AI Engineer to build out the agentic capabilities of Specter. This is primarily an engineering role – you'll be shipping production AI features.
We launched our AI Search Agent last year, and are now ready to take it much further. You'll be building the next generation of AI-powered experiences that help investors discover and research companies in ways that weren't possible before.
What you'll work on
Conversational AI interfaces within the product together with the frontend team
Tool use and function calling to expose Specter's capabilities to AI agents
Intelligent orchestration combining our proprietary data with web search and external sources
Deep search features that can find and synthesize very specific data across our data and web data
Evaluation frameworks to measure and improve agent performance
The underlying infrastructure to make these experiences fast and reliable
You'll sit within the backend team and work closely with our existing FastAPI codebase, extending it to support increasingly sophisticated AI workflows. The AI landscape moves fast – new models and frameworks emerge constantly – and we want someone who's excited to experiment, benchmark, and ship.
*Here are a few of the key people you'd be interacting with:
Henry** has architected our new Python backend and all client-facing APIs. With his vast experience he's a true unicorn, working across databases, frontend, backend, and infrastructure to make sure everything ties together nicely.
Josh leads the frontend team and has built and designed the majority of the product over the course of his close to 3 year tenure at Specter.
Jan both product manages the team and builds the first prototypes for many of our new features, such as network tracking or our AI agent. He brings experience as a serial founder in data and trading.
What we're looking for
Include the role's title in your subject line
Send links that best showcase the relevant things you've built and done
Please note that we are doing interviews on a rolling basis.
Interview process
Initial chat to discuss the role, learn about each other and explore whether it's a good fit
Coding exercise - a domain-specific exercise that doesn't require any preparation. No white-boarding!
One-hour general technical interview [+ exercise review]
Final interview - a chance to have a general discussion, meet the rest of the team and make sure we're the perfect fit for each other. You'll have learnt more about us by this stage, and this is a chance to have any remaining questions answered!