AI Field Engineer - Enterprise
$220,000 – $280,000 OTE (80/20 base/variable split). Variable component paid quarterly based on individual and team performance. Compensation scales with experience — candidates with 10+ years may be considered for above-range packages. Meaningful equity included on top of OTE.
We are looking for an AI Field Engineer (Enterprise) with 3+ years of experience to embed with Fireworks AI's most ambitious enterprise customers and turn complex GenAI challenges into production systems — fast. You'll be the technical tip of the spear, pairing deep hands-on engineering with the executive presence to earn trust across large organizations and drive deals from first discovery call to production deployment.
What will you be doing?
Lead technical discovery calls, scope POCs, and run load tests and evaluations to validate the right model architecture and deployment configuration for each enterprise customer
Build end-to-end POCs and production integrations hands-on-keyboard inside customer environments, navigating their infrastructure, security requirements, and organizational constraints
Guide customers on model selection, fine-tuning strategy (SFT, DPO, RFT), and evaluation frameworks — moving them from open-model exploration to production at scale
Manage multi-stakeholder enterprise relationships — identifying technical champions, navigating org politics, and aligning the right people to move deals forward quickly
Feed recurring customer pain points and deployment patterns back into the product roadmap, acting as a direct feedback loop between the field and engineering
Key Requirements
Deep hands-on experience with LLM inference and/or training — working knowledge of open-model frameworks (vLLM, SGLang, TensorRT-LLM) and fine-tuning workflows (SFT at minimum; DPO/RFT a strong plus); candidates with only closed-model/API-wrapper experience will not clear the bar
Proven ability to ship production code inside a customer's environment — not just advisory work; you've built and deployed POCs/MVPs that ran in someone else's prod system
Strong Python skills plus GPU/cloud infrastructure experience (AWS, Azure, or GCP) and comfort with Kubernetes
Executive presence and enterprise navigation skills — able to run a technical deep-dive with an ML engineer and present architecture trade-offs to a VP in the same afternoon
Pre-sales or customer-facing field engineering experience (FDE, Applied AI Engineer, Solutions Architect, or similar); pure software engineers without customer-facing exposure are not a fit
Fireworks is the fastest way to build, tune, and scale AI on open models. Ship production-ready AI in seconds on our globally distributed cloud infrastructure, optimized for your use case. Fireworks 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.
Team size
211 employeesFounded
Website
fireworks.aiTotal funding
Company locations
The AI Field Engineering team for AI Natives is a small, high-velocity unit led by Roberto Barroso Luque. At the time of this intake call, the team consisted of Roberto and two others — and the mandate is to grow fast by first bringing in a tight cohort of three to five exceptionally senior engineers who will become the leads of a much larger team within six months.
The culture is best described as flat, low-ego, and extreme-ownership-first. There are no senior, lead, or principal title distinctions — everyone is expected to roll up their sleeves, take ownership from day one, and do whatever needs to get done regardless of job boundaries. Arrogance is a non-starter; the team actively screens it out. As Roberto put it: "We don't vibe with people that are arrogant."
The AI Natives team operates at a distinctly faster pace than the Enterprise segment. The accounts they serve — companies like Harvey, Perplexity, and Sierra — are enterprise-scale companies that move at the speed of YC startups. That means engineers need to be comfortable with rapid context-switching, ambiguity, and learning on the fly. The expectation is that new hires are client-facing by week one or two, owning POCs and MVPs directly alongside account executives and customer engineering teams.
What sets this team apart from the Enterprise side is a deeper technical bar and a stronger product feedback loop. Engineers here are expected to push PRs directly to the Fireworks codebase, contribute to internal product initiatives, and translate field learnings into concrete platform improvements — not just advise customers. Product intuition and a genuine interest in shaping the roadmap are valued alongside engineering depth.
The team draws inspiration from forward-deployed and professional services models — think Palantir, BCG X, McKinsey Quantum Black — where engineers are both deeply technical and highly client-effective. Roberto himself came from BCG X and brings that structured, systems-thinking approach to how the team operates and how it hires.
Fireworks as a whole was founded by veterans of Meta PyTorch and Google Vertex AI, and that pedigree shapes the technical culture: rigorous, curious, and builder-first. The company embraces AI natively — candidates who haven't worked with open-source models or built GenAI features are not a fit. The team is looking for people who are hungry to learn, embrace AI as a craft, and want to build something genuinely new from the ground up.
Fireworks AI is a high-performance generative AI inference platform purpose-built for teams that need to ship production-ready AI fast. The platform enables developers and enterprises to build, fine-tune, and scale AI applications on open models — without the latency, cost, and lock-in that come with closed-model APIs.
At its core, Fireworks delivers industry-leading LLM inference speed on a globally distributed cloud infrastructure. Independently benchmarked as the fastest in the industry, Fireworks offers 15× faster speed, 4× lower latency, and 4× more concurrency compared to closed model alternatives — making it the infrastructure of choice for latency-sensitive, high-throughput production workloads.
Key platform capabilities include:
Optimized open-model inference — deploy and serve leading open-source LLMs at scale using frameworks like vLLM and SGLang, with support for quantization, optimal serving shapes, and GPU infrastructure tuning
Fireworks Training — a fine-tuning suite supporting SFT, DPO, and reinforcement fine-tuning (RFT) workflows, enabling teams to customize models for their specific use cases without sacrificing speed or quality
Evaluation frameworks — production-quality eval tooling that goes beyond benchmark scores to measure real-world performance metrics
Function calling and multimodal models — proprietary model innovations developed directly from Fireworks' production systems
Enterprise-grade infrastructure — SOC 2 compliance, data residency controls, and integrations with hyperscaler AI platforms including Microsoft Azure AI Foundry, AWS Bedrock/SageMaker, and GCP Vertex AI
Fireworks powers production workloads at companies like Uber, DoorDash, Notion, and Cursor — teams that have chosen Fireworks because their core products depend on AI running reliably, quickly, and at scale. Whether a customer is just beginning their open-model journey or already running open models in production, Fireworks provides the inference foundation and expert field support to get from prototype to production fast.