Senior Engineer

About Fuzzy

We build AI agents that handle the unglamorous parts of B2B sales. Finding the right prospects, commenting to stay on people's radar, replying in someone's actual voice, keeping the follow-up alive. All of it running in the background on LinkedIn while founders and sales teams get on with selling.

We're a small team based out of Singapore, working across time zones, and we're cashflow positive while we raise our seed round. Companies like MongoDB, Formlabs and Cyient already run on Fuzzy. For one of them we contracted 25 meetings and delivered 62. That's the bar.

The product is moving fast. We've shipped a memory layer that learns how a customer writes, a rebuilt unified inbox, and a set of agentic workflows that orchestrate campaigns end to end. There's a lot more half-built and a lot more we haven't started.

The role

You'd be a senior engineer with real ownership from day one. Small team means no layers and nobody to wait on. You pick up a problem, you ship it, you watch how customers use it, you fix what's wrong.

A lot of the interesting work sits where the app meets the model. You'll build features that lean on LLMs, then deal with everything that makes that hard in production: latency, cost, prompts that work in testing and fall over on real data, agents that need to make decisions without a human watching.

Some of what's on the table:

• Building and improving the agentic workflows that run campaigns, discover prospects, and draft replies
• Wiring up LLM-driven features across the full stack, from the prompt logic through to the UI a customer actually touches
• Making our message and reply quality better, which means getting properly into evals rather than eyeballing output
• Shipping product end-to-end with Next.js on the front and Node on the back

Who we're after

Two things matter more than any line on your CV.

• Curiosity. When the model does something weird, you want to know why before you bolt a retry on top. You read the docs and then go test the edges the docs don't mention. New tooling shows up and you've already had a poke at it by the time anyone asks.
• Resourcefulness. At our size nobody hands you a tidy spec. You're comfortable in the ambiguous bit, you find the workaround, you make a sensible call and keep moving. You'd rather ship something rough and learn than wait for perfect requirements that aren't coming.

If that's how you already work, you'll fit.

Requirements

• Full-stack, comfortable across Next.js and Node.js
• Confident working with AI IDEs day to day
• Familiar with AWS
• Hands-on experience implementing LLM APIs

Nice to have

• Experience with cloud agents
• Experience with prompt evals

The setup

Fully remote. We care about what you ship, not which chair you're sitting in. You'll work closely with a tight engineering and design team and have a direct line to the founders, because there isn't much between you and us.

If this sounds like your kind of mess to work on, get in touch and tell us about something you built when nobody told you how to.

Send in your application here: https://airtable.com/apppheevAU8BcwAMV/pagnrmnFP8nPEuXr9/form