Founding AI Engineer

Futureproofing is building an end-to-end AI platform for investment banking, and the near-complete MVP is ready for an expert to help it become a product real customers can rely on. This role is centered on production-grade AI engineering, where core product capabilities are built with LLMs, agents, RAG, and evaluation rather than added on after the fact.

As the founding AI engineer in Boston, MA (onsite), you will work closely with the founding team to define what gets built and why, then take ownership of the system areas that determine whether the platform performs reliably in the real world.

Responsibilities

• Take the product from MVP to something real customers rely on
• Build with LLMs, agents, RAG, and evaluation as core product surface areas rather than bolt-ons
• Partner directly with the founding team on scope and priorities, including rationale for engineering decisions
• Own high-impact parts of the system, including the data pipeline, infrastructure, and model integration
• Move quickly alongside engineers who already know how to ship

Requirements

• 7+ years shipping production software, including meaningful experience at an early-stage company
• Built at least one AI-powered product from nothing to production users, with the ability to explain key tradeoffs made during delivery
• Comfort across modern stacks, typically TypeScript/React/Node and/or Python, with experience spanning full-stack work into infrastructure
• Uses AI tooling daily as part of how you actually build
• Comfort with ambiguity and with being the person who decides
• Can communicate technical tradeoffs clearly to non-engineers without losing context for either side
• Founding engineer experience is a plus, noting the role does not include the traditional founding-engineer responsibility of building and owning the whole system solo

Technologies

• LLMs, agents, RAG, evaluation
• TypeScript, React, Node, Python

What Makes This Opportunity Different

• The product is category-defining
• Early stage, but not too early: the MVP is nearly built, so you join with something real in hand
• Build next to people who know the domain cold
• High-autonomy team