Fireworks AI — AI Field Engineer

Type: Full-time | Remote-friendly (US) with regular on-site client travel | New York, NY / San Mateo, CA / Remote (USA) Candidate compensation: $176K – $228K base (OTE $220K – $285K) + competitive equity Hiring count: 3 – 5 Visa sponsorship: Yes — H-1B transfers and TN visas sponsored; O-1 considered case-by-case Reports to: Roberto Barroso-Luque, Hiring Manager

About Fireworks AI

Fireworks is the fastest way to build, tune, and scale AI on open models — shipping production-ready AI in seconds on a globally distributed cloud infrastructure optimized per use case. It powers production workloads at companies like Uber, DoorDash, Notion, and Cursor, delivering 15× faster speed, 4× lower latency, and 4× more concurrency than closed models. Series C at a $4B valuation, backed by Benchmark, Sequoia, Lightspeed, Index, and Evantic.

Founded: 2021 | Team size: 181 | Total funding: $327M (incl. $230M Series C) Industry: AI, Software Development, API/SDK, Devtools, B2B, Enterprise Website: fireworks.ai Office: San Mateo, CA + New York, NY

Why Candidates Should Join

• Series C at a $4B valuation: $327M raised from Benchmark, Sequoia, Lightspeed, Index, and Evantic, with the $230M round freshly closed — well-capitalized with serious investor conviction.
• Real production traction at scale: Already powers workloads at Uber, DoorDash, Notion, and Cursor. Not a pre-revenue pitch — a platform running at enterprise scale now.
• Founding-team pedigree: Built by veterans of Meta PyTorch and Google Vertex AI — top-tier technical credibility in AI infra.
• Comp + equity: OTE $220K–$285K plus meaningful equity at a company still early enough for equity to matter.
• Own accounts from day one: Small team with a fast-growing mandate, a direct line to the product roadmap, and minimal bureaucracy.

Intake Call Summary

• Role: AI Field Engineer — a blend of sales engineering and forward-deployed engineering: integrating the Fireworks platform, client management, and performance engineering. The technical tip of the spear, embedded with the most ambitious AI-native customers.
• Team: Currently small; hiring 3–5 senior engineers to lead future growth.
• Must-haves: Engineering skills, experience building and maintaining systems, strong client management.
• Experience level: 5–10 years, ideally a senior IC ready to lead.
• Nice-to-haves: Kubernetes, ML engineering, product thinking.
• Culture: Fast-paced, low ego, extreme ownership.
• Comp/logistics: $220K–$285K OTE based on experience; equity available with high growth potential.
• Timeline/urgency: Immediate need to own accounts and POCs; goal is candidates client-facing within the first two weeks.
• Pain points: Current team is underwater with account work; needs top-quality, client-facing engineers.
• Ideal profile: Low ego, high ownership, hunger to learn; ML engineering over backend; fast-paced environment experience.

The Role

AI Field Engineers embed with Fireworks' most ambitious AI-native customers to turn complex AI problems into production systems, fast. The role sits at the intersection of engineering, product, and customer delivery — hands-on-keyboard building POCs, MVPs, and production integrations, while holding their own in executive-level conversations about architecture, strategy, and business outcomes.

What You'll Be Doing

• Build POCs, MVPs, and production integrations directly inside customer codebases.
• Own customer accounts end to end — manage relationships, run discovery, present to stakeholders.
• Do performance engineering and deployment on the Fireworks platform across customer environments.
• Channel client feedback into product improvements in collaboration with PMs.

Tech stack: Python, vLLM, SGLang, TensorRT-LLM, Kubernetes, AWS, Azure, GCP, Azure AI Foundry, AWS Bedrock, AWS SageMaker, GCP Vertex AI, LLM fine-tuning (SFT, DPO, RFT), GPU infrastructure

Qualifications

Seniority

• 5–10 years in customer-facing ML/AI engineering (FDE, Applied AI, or Solutions Engineering) as a senior IC — ~60% hands-on coding/deployment, ~40% client engagement + product feedback [Required]

Work Experience

• Built and shipped production ML/AI systems from the ground up [Must have]
• Direct client-facing engineering experience — ran POCs/MVPs, managed accounts, presented to stakeholders [Must have]
• Archetype A: FDE/embedded engineer at AI-native startups or professional services firms — consulting/delivery focus [Required]
• Archetype B: Senior ML/AI engineer with training, fine-tuning, inference, or model deployment, plus demonstrated client-facing experience [Required]
• Background at an AI/ML startup/infra company, professional services firm, or big tech with client-facing exposure [Strongly preferred]

Hard Skills

• Strong Python + hands-on ML/AI engineering depth [Must have]
• Experience with inference serving frameworks (vLLM, SGLang) and cloud GPU infrastructure (AWS, GCP, Azure) [Strongly preferred]
• Familiarity with Kubernetes [Strongly preferred]

Soft Skills

• Collaborated with PMs to rapidly channel client feedback into product improvements [Must have]
• Low ego, extreme ownership [Required]

Miscellaneous

• Comfortable with regular on-site customer visits (AI-native accounts move at YC startup pace) [Required]

Traits to Avoid

• Pure advisory/SA profiles who have never shipped code in a customer's production environment
• No AI/ML exposure on their resume — no open-source model experience, no GenAI features built or integrated
• Multiple job tenures under 1 year
• Pure Big Tech IC without any external exposure

Role Details

Candidate salary$176K–$228K base (OTE $220K–$285K)EquityCompetitive equityOn-site policyUS-based; open to remote or in-office in New York, NY or San Mateo, CA; regular on-site customer travel expectedVisa sponsorshipH-1B transfers and TN visas sponsored; O-1 case-by-caseEmployment typeFull-timeLocationNew York, NY / San Mateo, CA / Remote, USA

Screening Questions

1. Are you based in the US and comfortable with regular on-site travel to client offices?
2. Describe a time you built and shipped a POC or production integration directly inside a customer's codebase — what was the stack and what did you deliver?
3. What ML/AI systems have you built from the ground up? Walk me through the architecture and your specific contributions.
4. How have you taken client feedback and channeled it into product improvements internally?
5. What's your comfort level with ambiguity and fast context-switching across multiple accounts?
6. What is their salary expectation?
7. How actively is this candidate exploring new opportunities?

Interview Process

Stage 1 — Submit candidate After submitting, you're notified if the hiring manager wants to proceed.

Stage 2 — Take-Home Assignment (Self-paced) Build a working text-to-SQL system, submitted async.

Stage 3 — Recruiter Screen (30 min) Conversation: logistics, motivation, role fit.

Stage 4 — Discovery + Hiring Manager (45 min) Live discovery role-play with the hiring manager.

Stage 5 — Culture + Live Coding (1 hour) Product and production discussion, then extend your take-home code live.

Stage 6 — On-Site Final Loop (~2 hours) Customer demo/presentation + executive values conversation.

Stage 7 — Executive Interview (30 min)

Stage 8 — Debrief (60 min)

Stage 9 — Pre-Offer (60 min) Auto-approved for all future roles with this client.

Stage 10 — Offer Extended

Stage 11 — Candidate Hired

Ideal Companies & Backgrounds

Updated Jun 22, 2026

For sourcing reference — these companies and adjacent companies are good starting points.

Ideal Companies Harvey, C3 AI

Forward-deployed / professional services engineering firms (explicitly mentioned by HM) Palantir Technologies, BCG X, C3.ai Digital Transformation Institute, Scale AI

AI-native startups and direct competitors with FDE or embedded engineering motions (explicitly mentioned by HM) Together AI, Baseten, Anyscale, Modal Labs, Replicated, Groq, Cohere, Perplexity AI, Harvey AI, Sierra Nevada Corporation

Big Tech with strong ML/AI engineering and some client-facing exposure (explicitly mentioned by HM) Google DeepMind, Meta AI, OpenAI, Anthropic, NVIDIA, Databricks, Snowflake, Hugging Face

AI-native inference, MLOps, and LLM infrastructure companies (highest-priority talent pool — deep open-model and serving framework experience) Together AI, Replicated, Modal Labs, Baseten, Anyscale, OctoAI, Groq, Cerebras, Mistral, Cohere

Hyperscaler AI platforms and cloud infrastructure with LLM/GPU deployment experience (Azure AI, AWS, GCP — in JD and intake) Microsoft, Google, Amazon Web Services (AWS), NVIDIA, AMD, Databricks, Snowflake, MongoDB

AI-native developer tools and production AI application companies (hands-on LLM integration + customer-facing field engineering) Cursor, Notion, Scale AI, Weights & Biases, Hugging Face, LangChain, Pinecone, Weaviate, Glean, Perplexity AI

Non-Ideal Companies (avoid sourcing)

Hyperscaler cloud solutions architects (HM: better fit for Enterprise role, not AI Natives) Amazon Web Services (AWS), Microsoft Azure DevOps

Pure closed-model API wrapper companies — engineers only work with OpenAI/Anthropic APIs, no open-model inference or fine-tuning (flagged disqualifying by David in intake) OpenAI, Anthropic, Jasper, Copy.ai, WRITER, Typeface

Traditional enterprise SaaS where AI is a bolt-on feature layer (candidates lack AI-native depth and open-model experience) Salesforce, ServiceNow, Workday Peakon Employee Voice, SAP, Oracle, HubSpot, Zendesk

Ideal Candidate Profiles

For reference only — do not source these specific profiles.

Anthony Nguyen — LinkedIn AI Engineer | Coral Springs, United States

• Strong companies (C3 AI)
• C3 is a professional services firm — likely some client-facing exposure
• Worked with AI
• Areas for improvement: nothing on resume explicitly shows client-facing experience ("none of it seems to be client facing"); Ravi flagged him as looking like "a backend engineer"

Ameer Qamar — LinkedIn Building AI Voice Agents for CX | Toronto, Canada

• Resume was specific — tied buzzwords to actual projects/company work, not generic
• AI voice agent experience — relevant AI domain
• Cresta — a Fireworks client, so familiar with the context
• University of Waterloo — strong technical bar
• Areas for improvement: "Would interview him, but he's not top top. He's a pass, let's move it forward."

Rejected Candidate Feedback

• None recorded yet.