Asks for

The ad says“Familiarity with agent search patterns: agent file search, vector RAG, vectorless RAG”
The ad says“Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine”
The ad says“Prompt and context engineering experience (e.g., understand the nature of language model responses, able to return equal results out of smaller models with better prompts and context)”
The ad says“Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine”
The ad says“Full-stack development with a stronger spike in backend (Python is core; TypeScript helpful)”
The ad says“Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine”
The ad says“Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine”
The ad says“Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine”
The ad says“Process orchestration and workflow automation experience”
The ad says“Prompt and context engineering experience (e.g., understand the nature of language model responses, able to return equal results out of smaller models with better prompts and context)”
The ad says“Full-stack development with a stronger spike in backend (Python is core; TypeScript helpful)”
The ad says“Familiarity with agent search patterns: agent file search, vector RAG, vectorless RAG”
The ad says“Familiarity with agent search patterns: agent file search, vector RAG, vectorless RAG”
The ad says“Process orchestration and workflow automation experience”

The role

This is the FDE role with an applied AI spike

You'll be the engineer on a client engagement, working on the Titan platform and toolkit to deliver agentic systems that solve real banking workflows. You won't be alone — you'll be supported by senior FDEs and the platform team — but you'll own the build. You'll spend most of your time inside Titan Foundry, our Banking-Reasoning Models, and our Banking Agents framework, configuring and extending to fit each client's specific problem.

The growth path is into the senior FDE archetype: more off-platform building, deeper partnership with client engineering teams, broader architectural ownership.

What you'll do

Build and ship agentic workflows on the Titan platform to solve specific banking use cases (lending, credit, compliance, ops)
Configure agents, write integrations, design evals, and wire up the observability that proves things work in production
Partner directly with clients — operators, analysts, sometimes their engineers — to gather requirements, demo progress, and absorb feedback
Iterate fast: hypothesis, prototype, feedback, refine, ship
Capture patterns back into the platform so your next build (and everyone else's) starts ahead
Learn the banking domain rapidly through immersion — risk, controls, lending lifecycle, regulatory context

Who you are

A builder. Demonstrably. We'd rather see a GitHub full of unfinished experiments than a resume full of polished titles.
Unconventional paths welcome — and frankly preferred. If you taught yourself to code while working in a different industry because you couldn't stop thinking about the opportunity, we want to talk.
Comfortable being client-facing or excited to become so. You don't need years of consulting; you do need to want to be in the room.
Hungry to learn banking. You don't need to walk in fluent — you need to walk in curious.
Capable of running a small project end-to-end without dropping balls.

Technical foundation

Prompt and context engineering experience (e.g., understand the nature of language model responses, able to return equal results out of smaller models with better prompts and context)
Full-stack development with a stronger spike in backend (Python is core; TypeScript helpful)
Experience building agents with LangChain, LangGraph, DeepAgents, and model-provider SDKs (Claude SDK, OpenAI Agents SDK) — strong side-project experience is fine
Familiarity with agent search patterns: agent file search, vector RAG, vectorless RAG
Some exposure to AI evals, observability, or training loops — you've at least played with the tooling
Awareness of AI security concerns (data controls, gateways, observability) and willingness to learn the rest
Process orchestration and workflow automation experience

We don't expect a perfect match on every line. We expect that what you don't yet know, you can pick up in weeks, not quarters.

Why Titan

We're building the AI platform for banking because most of the two million people employed by 4,000+ U.S. banks cannot safely and securely use modern AI tools at work — and none of them are using AI purpose-built for their domain. We're changing that. Our platform includes:

Titan Foundry — secure gateway to top AI models for bank employees
Banking-Reasoning Models — our flagship small language models for financial services, trained by ex-regulators and banking leaders
Banking Agents — a configurable AI application layer with expertise in lending, compliance, underwriting, and operations

Our team comes from Goldman Sachs, KeyBank, and the Office of the Comptroller of the Currency. We've built and sold companies into the top 25 U.S. banks. We're pre-Series A, small, and shipping.

The path here is: ship on the platform → master the toolkit → grow into off-platform builds and senior architectural ownership. We're not hiring you to stay where you start.

What we offer

Competitive base salary plus meaningful pre-Series A equity
Remote-first with regular client travel (~25–40%, depending on engagement)
Unlimited PTO
100% of employee medical, dental, and vision premiums covered by Titan