We're building Cursor for RFPs.

The way companies compete for work is being rebuilt, and Grizzly is building the platform at the centre of it.

Governments in NZ and Australia spend over $150 billion a year awarding contracts through competitive bids. Add the private sector and that number multiplies. Every one of those contracts starts with companies responding to formal requests — RFPs, RFTs, RFQs, ROIs — and the process is painful: thousands of pages to understand, evidence to gather from across the business, and a polished response to compile, all under a hard deadline. For every winner, there are several others who spent hundreds of thousands of dollars on a response that was unsuccessful.

Grizzly is building an AI-first product that helps teams put their best foot forward in competitive bids, fast — combining a company's institutional knowledge with AI agents that analyse, draft, and refine bid responses within a single workspace.

We're a privately funded, five-person team based in Auckland. We've closed customers in the construction and engineering verticals, with strong traction in our pilots and no shortage of interest. We're moving fast, and looking for someone who wants to build with us.

About the Role

This is a hands-on, AI-heavy role. If AI, harness design, context engineering, and making agents work reliably in production sound like interesting areas, then this is your kind of role.

As engineering lead, you'll own how Grizzly gets built - the architecture, tooling, and technical direction behind the product. Most of your time will be spent in the codebase - shipping features, designing systems, and working on the agent harness that our customers rely on.

You'll report directly to the CEO and collaborate closely with our product lead to shape what gets built and how. This is a flat, high-agency team. We value autonomy, collaboration, and impact over hierarchy or rigid role boundaries.

In this role, you will:

• Own the technical side of the product end-to-end - architecture, infrastructure, security, and the decisions behind them

• Evolve the agent harness - what the model sees, when it sees it, and what it can act on

• Improve and extend the product - new features, bug fixes, and keeping the codebase healthy as it grows

• Shape what gets built - work closely with product to decide what's worth solving, how to approach it, and when it's good enough to ship

• Contribute to a culture of high ownership, low ego, and thoughtful collaboration.

You might thrive in this role if you:

• Have 5+ years of experience building and shipping software in production end-to-end

• Are comfortable across the full stack - React, TypeScript, AWS or similar - and can move across the full stack as needed

• Have an interest in AI, LLMs, and agent harness design (direct experience not required)

• Care about why you're building something, not just how - and are comfortable with scope that's defined collaboratively

• Work well without a playbook - you've operated in environments where the path forward wasn't obvious and found that energising

• Know the difference between code that needs to last and code that’s needed to learn - and are comfortable writing both

• Bias toward trying things - you'd rather test an idea quickly than debate whether it'll work

• Are a genuine collaborator - you think out loud, raise concerns early, and commit when a decision is made

• Care deeply about customers and get energy from delivering something they love

Bonus if you have:

•  Experience at a 0-to-1 startup

•  Hands-on experience building AI-driven products

•  Familiarity with SST, NoSQL databases, Vercel AI SDK

•  Experience with agentic coding tools like Claude Code, Codex, or Cursor

•  Knowledge of ISO 27001, SOC2 or equivalent security frameworks

•  Experience with RFPs, tenders, or competitive bid scenarios

Why our team loves working at Grizzly

Work that matters: The problem is real, the market is massive, and what you build will have a direct, visible impact on the people who use it.

Real autonomy: You'll have the freedom to experiment, try new approaches, and shape how things are done, both in the product and how we work.

AI-first culture: We embrace new tools and push the boundaries of what they make possible - it makes us faster, and what we learn feeds directly back into the product we build for customers.

A friendly team that takes the work seriously but themselves lightly - this isn't easy, and we wouldn't have it any other way. Good humour and strong relationships are what make hard work worth doing.

Competitive compensation: This role has a range of $180k – $200k per year, with both your scope and compensation growing as Grizzly does

Hybrid setup: three days in the office, two from home, with genuine flexibility when life happens

Unlimited tokens for use with frontier models: As well as budget for any tool or resource that supports your work or learning.