Founding AI Engineer (Agent Systems)

Founding AI Engineer, Agent Systems

Boston or NYC · Hybrid · Full-time · $160K - $190K + equity

Life sciences sales reps operate in unusually complex environments, supporting surgeons, physicians, and clinical teams across pharma and medical device. They navigate clinical evidence, competition, patient access, and tightly regulated information, often making decisions about what to say or the best actions to take while alone in the field without support or the ability to record conversations. When reps struggle to find the right information, recognize an issue, or get the right support, the consequences extend beyond sales performance. A clinician may not get the treatment information they need when they need it, a surgeon may not feel confident using a device, or an important problem in the field may never make its way back to the organization.

Vincer is building a voice-first AI decision support system for these field teams when judgment, scientific knowledge, and speed matter most. The product has to understand messy, unstructured situations, assemble the right context, retrieve approved knowledge, orchestrate multiple AI systems, know when to escalate, and respond fast enough to feel conversational.

We're looking for a founding AI engineer who wants to own the intelligence layer of a voice-first AI system and shape how we build AI for our life sciences customers. You'll be one of the first two full-time engineers, joining alongside a Founding Software Engineer and the co-founders and technical advisors. There is no engineering organization above you handing down decisions. You'll inherit multi-agent workflows, existing evals work, and a working MVP, and get to decide what to keep and set the AI direction and future roadmap with us.

THE ROLE

You will own Vincer's intelligence layer: agent architecture and orchestration, evaluation, retrieval and grounding, guardrails, and the AI quality that determines whether end users and enterprise buyers trust the product.

The system is agentic by design: it decides when to act, not only what to say. Agents work from rep activity and accumulated context rather than waiting to be prompted, coordinate across multiple models and data sources, and make a judgment call about whether to handle something directly, route it, or escalate it to a human.

The hard problem is not getting that to work once. It is making it hold up: measurable, predictable, and safe across foundation models, customer knowledge, and proprietary data, in a regulated environment where a wrong answer has legal consequences beyond a bad user experience.

Examples of Core Projects

• Evolve the orchestration layer so specialized agents can be independently versioned, deployed, and improved

• Build the evaluation systems that make agent quality visible: failure modes, regressions caught before customers see them, and honest answers about whether a change helped

• Build retrieval and grounding over approved customer content, where unsupported answers are treated as system failures

• Own guardrails, escalation logic, and safety behavior for a regulated environment, with guidance from advisors who have implemented AI systems inside life sciences organizations

• Engineer for latency and cost through context management, caching, model selection, and routing across model tiers

• Encode expert judgment into system behavior, working directly with our scientific co-founder

WHO WE NEED

Our ideal candidate has worked in a startup environment in recent years and wants to join a new venture at its earliest stages. You understand that early product development requires fast iteration across AI, platform, systems, and customer learning. You are located in Boston or NYC and understand that regular, in-person collaboration is essential to build the best products.

Required Experience

• 5+ years of full-time software engineering, including at least 2 years of recent, hands-on experience building and shipping AI systems that you owned in production, not prototypes or demos

• Experience designing or working with agentic systems: routing, tool use, retrieval, context management, and multi-step workflows

• You've built across multiple model providers and have views on where the current agent tooling helps and where it gets in the way

What Would Make You Stand Out

• Enterprise software delivery, especially in a regulated industry (healthcare, finance, or similar)

• You've been on the hook for an AI system after launch, not just through launch

• Comfort encoding expert judgment into measurable system behavior, working alongside domain experts

• Experience with voice or other latency-sensitive interfaces

• Rigor about measurement. You would rather have a smaller improvement you can prove than a larger one you cannot, and you understand the strengths and limits of model-based evaluation

How You Work

You care about the code you leave behind. An agent system that works in a notebook and an agent system that survives in production are different artifacts, and you know the difference. You use modern AI coding tools as a matter of course and expect a small team to ship beyond its size. You want broad ownership, fast decisions, and the accountability that comes with them.

The intelligence layer is where Vincer earns trust. We're looking for the engineer who wants to own it. If that’s the kind of challenge you're seeking, we’d like to hear from you!

***At this time, Vincer is unable to hire candidates who require employment visa sponsorship or employer participation in work authorization programs.***