Careers

Build AI the public sector can trust.

We build systems that governments and regulated institutions rely on. The bar is high because the consequences are real — and so is the satisfaction of shipping something that holds.


How we work

Four habits. Not values on a wall — the things we check for in review.

  • 01

    The work is the proof

    Our buyers verify everything. So do we — claims in code review, in evaluations and in what we say to clients.

  • 02

    Boring where it counts

    Novel models, conventional infrastructure. Reliability is a feature the client cannot see until it is missing.

  • 03

    Write it down

    Decisions are recorded as they are made. The next engineer should never have to reverse-engineer why.

  • 04

    Own it to the end

    The people who design a system help run it, and stay until the client’s team can.


Teams

Where you would fit. Small teams with end-to-end ownership, working directly with the client.

  • Applied machine learning

    Fine-tuning, retrieval, evaluation and red-teaming on client data, in client environments.

  • Platform & infrastructure

    GPU clusters, inference serving and MLOps across cloud, on-premise and air-gapped estates.

  • Product engineering

    Full-stack engineers building the copilots, search and platforms people use every day.

  • Security & compliance

    Threat modelling, assurance and the evidence that lets a client say yes.

  • Delivery

    Engagement leads who scope the work with clients and keep it on course through procurement and delivery.

  • Public-sector advisory

    Former public servants who help clients through impact assessments, procurement and policy.


Where we work

  • CA

    Canada

    Client-facing delivery, security and advisory. Some roles require eligibility for a Government of Canada security clearance.

  • IN

    India

    Engineering and machine-learning teams working alongside Canadian delivery leads, with overlapping hours by design.


How we hire

  • 01

    Introduction

    A 30-minute conversation about what you have built and what you want to work on next.

  • 02

    Technical session

    A working session on a realistic problem from our domain — no puzzles, no take-home marathon.

  • 03

    Team conversations

    Meet the people you would work with, and ask them anything.

  • 04

    Decision

    A decision within a week of the final conversation, with feedback either way.


No open roles are listed right now. We post them here when they open — and we read every speculative application.

Send what you have built and the team it fits. A link to the work beats a CV.