Who are Fireworks AI's decision-makers?

Lin Qiao is CEO and co-founder, and the primary public face of the company. She holds a PhD in Computer Science from UC Santa Barbara and spent over a decade at Meta, most recently as Head of PyTorch leading 300+ engineers. Under her leadership, Meta's PyTorch served 5 trillion inferences per day — a direct proof-of-concept for the scale Fireworks is now commercializing. Prior to Meta she was a tech lead at LinkedIn and an engineer at IBM, giving her both infrastructure engineering depth and enterprise go-to-market instincts.

Dmytro Dzhulgakov (CTO, co-founder) was a PyTorch core maintainer at Meta and is the principal architect of Fireworks' FireAttention CUDA kernels and FireOptimizer inference orchestration system. The other five co-founders — Benny Chen (Meta ads infrastructure lead), Chenyu Zhao (Google Vertex AI Lead; also has background at NVIDIA and UC Berkeley), James Reed (PyTorch compiler, Meta), Dmytro Ivchenko (PyTorch ranking lead, Meta), and Pawel Garbacki (Meta Newsfeed ML lead) — hold engineering and product leadership roles that blend deep ML systems expertise with practical experience running inference at massive scale.

On the GTM side, Marc Holmes joined as CMO (circa 2024) and Bardia Shahali joined as VP of Sales, with Shahali credited with driving 4x ARR growth across North America. Their addition reflects Fireworks' transition from purely bottoms-up developer-led growth to a hybrid model with structured enterprise sales motion layered on top of self-serve.

Fireworks AI has a split buying committee that mirrors its bottoms-up commercial motion. For developer-led workloads under approximately $50K/year, engineers and individual developers self-serve via credit card — no sales involvement required, and spending tiers expand automatically as usage history is established. This frictionless entry point is intentional: the company converts PLG usage into enterprise contracts as workloads scale.

For mid-market and enterprise contracts, the primary economic buyer is typically a VP of Engineering, Head of AI/ML Infrastructure, or Director of Platform Engineering, who owns the inference budget and drives vendor evaluation. Technical sign-off typically comes from a Staff or Principal Engineer who has run benchmarks on FireAttention vs. competing solutions. At the C-suite level, Lin Qiao and the founding team make strategic partnership decisions — model provider integrations, cloud marketplace deals with AWS and GCP, and hardware partnerships with NVIDIA and AMD.

Bardia Shahali's sales team owns expansion from PLG accounts into enterprise contracts. Fireworks is listed on both AWS Marketplace and Google Cloud Marketplace, enabling enterprises to apply existing cloud commit spend and dramatically compressing the procurement timeline. For sellers pursuing Fireworks as a prospect, the enterprise sales and revenue operations team (actively hiring AEs and RevOps in 2026) is the right target for vendor conversations; the engineering organization is the right target for technical integration partnerships.

Fireworks operates as a lean, engineering-heavy organization with approximately 150 employees as of mid-2026, implying revenue per employee of roughly $5M+ annualized — an unusually high figure that reflects the infrastructure company's automated delivery model and the leverage of the proprietary FireAttention and FireOptimizer stack over commodity serving solutions. The founding team of seven appears to have retained significant influence over technical direction; CTO Dmytro Dzhulgakov is still actively publishing research and architecture decisions publicly.

As the company approached and crossed $300M ARR, it began investing in GTM structure: a CMO, VP Sales, and SVP Finance were all added in 2024–2025. Active job postings in 2026 show hiring in enterprise account management, revenue accounting (ASC 606 compliance for SaaS and usage-based revenue), GPU capacity forecasting, and marketing operations — all signals of a company building the organizational infrastructure to support either a large enterprise scale-up or a future IPO process.

Fireworks' engineering culture is visibly open: the company publishes detailed technical blog posts on FireAttention kernels, FireOptimizer architecture, and benchmark methodology, and maintains an active Discord community as a primary developer support and feedback channel. This openness is a deliberate recruiting and developer-trust strategy, consistent with the founding team's background in the open-source PyTorch ecosystem.