Technical Product Manager

IMPORTANT: This isn’t a traditional product manager role. If your first instinct when you have a question is to ask an engineer, this isn’t for you. If you measure yourself by features shipped or ceremonies run, this isn’t for you. But if you’ve built your own AI agent to scratch a personal itch, shipped a side project using Lovable or Cursor just to see how far you could take an idea, or read a product spec and immediately rebuilt a prototype before anyone asked you to, then we should talk.

About Dragonfly: Have you ever wanted to build something that doesn’t just improve the status quo, but makes it 100x better? Not just a small step forward, but a complete reinvention. That’s what we’re doing, and we need the best people to help us make it happen. We’re an early-stage start-up out of London working at the intersection of system design, cloud infrastructure and AI to accelerate the adoption of modern tools and technologies. Our goal is nothing less than to enable businesses of all sizes and shapes to leverage best-of-class technologies and latest advancements in GenAI, including agentic frameworks. We’ve recently launched our product, having raised a hefty pre‑seed fund from Tier 1 investors with additional contributions from some awesome angels to bring the vision to life. We are now in delivery mode and have already partnered up with 20+ world‑renowned brands to bring the product to the market. Sounds exciting? This is a rare opportunity to join at the earliest stage and help build the foundations of a generational product.

What you’ll be doing:

• Focusing on the "slow things" that let the team move fast: the decisions, definitions, and direction that engineers shouldn't need to stop for.
• Participating in two‑week sprint cycles, with a target of one‑week sprints: thinking one sprint ahead while the current one is in flight, enabling fast feedback loops without pulling engineers into every conversation.
• Leading strategic sessions: clarifying what's been decided, preventing the team from revisiting settled questions, and defending product direction under pressure.
• Defining what good looks like from a user's perspective: if the product suggests Datadog alternatives, you know whether that list would make a senior buyer trust it or dismiss it, and you use that judgment to shape what we build.
• Shaping how our intelligence agents reason: what they should weigh, surface, and present.
• Being the team's deepest source of user understanding: through genuine ongoing proximity to the people who buy and use these products, not periodic research sprints.
• Working with a generative product surface: the product composes interfaces from data and AI on demand; you define what users need to understand, not what the screen should look like.
• Helping set the foundational design guardrails the team can build within: working with our designer to move from pixel‑level decisions to principles that scale.
• Staying at the frontier of AI‑powered product development: trialling new tools and techniques, evangelising what works, and raising the bar on what fast looks like. This is a core responsibility, not a nice‑to‑have.

We are looking for someone who:

• Has a technical background, ideally a degree in Computer Science, Engineering, or equivalent.
• Fully fluent in how software is built: you understand what's happening under the hood and never create noise by asking what you could find yourself.
• Reaches for internal documentation, agents, and tooling first. Self‑reliance isn’t a preference here, it’s the baseline.
• Is AI‑native, not occasionally, but as default. Claude Code, Cursor, Lovable, and similar tools are how you work. You build prototypes before meetings, validate hypotheses without pulling in engineering, and move faster because of it. If you're not already using AI tooling daily to ship real work, this isn't the right role.
• Has already taken a product idea as far as it will go on their own, built a prototype, shipped a side project, entered a hackathon (without waiting to be asked).
• Thinks in components and generative interfaces, not fixed screens and wireframes.
• Can look at a product comparison and immediately articulate what’s useful, what’s missing, and what would make a senior buyer trust it.
• Makes and defends product direction decisions: doesn’t just synthesise, but takes a position and holds it.
• Is comfortable moving fast with imperfect information; knows validation is cheap and acts accordingly.

Our stack: You won’t code day‑to‑day, but you need to understand what’s happening under the hood and be comfortable working directly alongside the systems that power the product. Frontend TypeScript · Next.js + React (App Router, RSC) · Shadcn UI, Radix UI.