AI Product Engineer

New York, NY;
Remote, USA;
San Mateo, CA

At Fireworks, we’re building the future of generative AI infrastructure. Our platform delivers the highest-quality models with the fastest and most scalable inference in the industry. We’ve been independently benchmarked as the leader in LLM inference speed and are driving cutting‑edge innovation through projects like our own function calling and multimodal models. Fireworks is a Series C company valued at $4 billion and backed by top investors including Benchmark, Sequoia, Lightspeed, Index, and Evantic.

We’re an ambitious, collaborative team of builders, founded by veterans of Meta PyTorch and Google Vertex AI.

Product engineers at Fireworks build the software we ship to our users. You'll join a high‑autonomy team that turns frontier AI infrastructure into products developers and enterprises love — from the inference API and developer console to fine‑tuning workflows, deployment management, billing, and new bets like Fire Pass and agentic tooling.

This is not a "take a ticket and implement" role. You'll sit close to customers, talk to users directly, and iterate on what you ship based on real feedback — often within days, not quarters. You'll own features end‑to‑end: understand the problem (sometimes by joining a customer call), design the solution, build it across the stack, ship it, instrument it, and improve it until it works in production.

Fireworks owns the full stack: training, fine‑tuning (SFT and RFT at frontier scale), serverless and dedicated inference, multi‑LoRA serving, and agentic orchestration — all on infrastructure we built from scratch. Product Engineering sits on top of that stack and makes it usable. The pace of AI moves week to week; the map is being drawn as we go, and product engineers here help draw it.

You may be a fit if:

• You've built a product people actually use — and you cared about how it felt, not just whether it compiled
• You blend strong engineering with product taste and an instinct for what great developer experience looks like
• You'd rather talk to a customer, ship a prototype, and learn than wait for a perfect spec
• You have a knack for making powerful infrastructure feel simple — without dumbing it down
• You move fast with good judgment: you know when to polish and when to ship and iterate

____________________________________

• Ship product and talk to users. Join customer calls, work with design partners, and turn what you learn into product improvements — not slide decks
• Own features end‑to‑end across the Fireworks platform: developer console, API surfaces, model playground, fine‑tuning workflows, deployment management, billing, and new product bets
• Sprint on 0→1 initiatives. Take ambiguous ideas from concept to shipped product in weeks — new workflows, new product surfaces, new ways for developers to build on Fireworks
• Build for deeply technical users where the product is an API as much as a UI — and sweat the details in both
• Work full stack across frontend (React/Type Script), backend services (Python, Go), APIs, and data layers to deliver polished, production‑quality surfaces
• Iterate with data and conviction. Instrument what you ship, watch how developers use it, run experiments, and improve fast
• Partner across the company with Infrastructure, Applied Research, Growth, Data, and Design — but don't wait for perfect alignment to make progress
• Punch through technical walls. When platform primitives aren’t there yet, you'll help define what's needed and build around (or through) the gaps

____________________________________

• Launching a new developer workflow end‑to‑end in a two‑week sprint — from customer problem to shipped console + API surface
• Redesigning a core product flow (deployments, fine‑tuning, or billing) after sitting in on customer onboarding sessions
• Building and iterating on agentic product surfaces — tool use, orchestration UX, and harness tooling that developers trust
• Running experiments on API/console behavior to reduce time‑to‑first‑successful‑request for new developers
• Shipping a new consumption model or product vertical from scratch with a small cross‑functional pod

____…