Role Overview

We're looking for a strong Full Stack Engineer to help build the products that put AI into the hands of farmers and agricultural operators - as well as the internal tools our team relies on every day. Our stack is TypeScript/Node and React, so you'll either be comfortable there already or be a strong enough engineer to ramp up quickly. What matters most is that you can build and ship real features that real users depend on.

The work involves building on top of AI and agent frameworks (like LangGraph), but you don't need to be an AI expert on day one. If you're a capable engineer who ships clean, working software, is genuinely curious about AI, and wants to learn how to build with agents in production - you'll do well here.

This role sits alongside our Infrastructure & DevOps Engineer on the core software team. You'll contribute to either our internal AI platform (the Velocity Engine) or our external-facing AI products, depending on team priorities.

This is a hands-on building role in an early-stage startup. You should be comfortable working with ambiguity, limited documentation, and light process - and be motivated by shipping over perfecting.

This role is hybrid working out of our Southeast Calgary office 3 days per week.

Key Responsibilities

Build The Product

• Design and build full stack features across our web frontend (React) and backend (Node/TypeScript, Fastify, Prisma).
• Integrate LLM and agent frameworks (LangGraph and similar) into product features - tool use, multi-step workflows, streaming responses.
• Turn rough requirements into working, shipped features, iterating quickly based on feedback.

Ship & Operate

• Take features from development through staging to production; keep them reliable and observable.
• Respond to issues, debug problems (including the non-deterministic quirks of LLM-driven features), and improve the system as you go.
• Work with the Infrastructure & DevOps Engineer to keep deployments smooth and scalable.

Collaborate

• Work closely with product and AI/ML teammates on what to build and how it should behave.
• Add tests, monitoring, and evaluation for the features you own so we can trust them in production.

Relevant Experience

• Solid full stack engineering fundamentals - you can build and ship a feature end to end.
• Strong programming skills in a modern language, and experience building web applications (frontend + backend).
• Comfortable working in TypeScript/Node and React, or confident you can pick them up quickly (this is our stack).
• Experience designing and consuming APIs in real applications.
• Genuine curiosity about AI and agents, and eagerness to build with them.
• Comfortable in a fast-moving startup with ambiguity and light process; a bias for shipping.

Preferred Qualifications

• Any hands-on experience with LLMs or agent frameworks (LangChain, LangGraph, CrewAI, etc.) - even side projects or prototypes.
• Familiarity with prompt engineering, evaluation, or monitoring of AI features.
• Python familiarity (some of our tooling touches it).
• Experience with cloud platforms (AWS preferred) and Docker.
• Exposure to agtech, robotics, autonomous systems, or developer platforms.
• Open-source contributions that show ownership.

What You Bring

• Bias for shipping - you build working software, not just demos, and see it through to production.
• Curiosity - excited to learn AI/agent tooling and figure things out without a lot of hand-holding.
• Ownership - you take responsibility for what you build, from first commit through deployment.
• Breadth over specialization - you enjoy jumping between frontend, backend, and product problems.

Not Sure You Meet Every Requirement?

Research shows that some candidates, especially women, underrepresented groups, and career changers, are less likely to apply for a role unless they meet 100% of the listed qualifications. At Precision AI, we believe the right person can grow into the role, and we value potential as much as experience. If you're excited about our mission and think you could contribute, we encourage you to apply, even if you don't check every single box.