Founding AI Engineer (Agentic AI + Reinforcement Learning)
Stealth AI Startup – AI for Semiconductor Engineering
San Francisco – Onsite
$180k – $250k base + 0.4–0.75% equity
Early Team | Real Production Customers
This isn’t another “LLM wrapper” job.
We’re working with a stealth AI startup building agentic AI systems that automate complex chip verification and debugging workflows.
Today, debugging a single issue in chip verification can take days and multiple engineers digging through tools, logs, and simulation environments.
This company is building AI agents that reduce those workflows to minutes.
The platform captures how the best silicon engineers solve problems and turns that knowledge into AI playbooks that scale across entire engineering teams.
Customers already include fast-moving semiconductor startups and large publicly traded chip companies.
They’re now hiring a Founding AI Engineer to help build the core intelligence layer powering these agents.
This is a true 0 → 1 role where you’ll design systems, ship production models, and directly shape the future of AI in hardware engineering.
What You’ll Own
Build the Intelligence Layer
Design and implement post-training pipelines that shape how AI agents reason about chip verification workflows.
Reinforcement Learning Systems
Develop reward models, RLHF / RLAIF pipelines, and evaluation harnesses that improve agent decision-making over time.
Agent Tooling & Execution Systems
Build the infrastructure that allows models to interact with real engineering environments and EDA tools.
Production AI Systems
Ship agentic AI systems used directly by semiconductor engineers solving real verification challenges.
Customer-Driven Intelligence
Work closely with silicon engineers and customers to encode domain knowledge into training data, reward signals, and evaluation frameworks.
Frontier AI Work
Stay close to advances in LLM reasoning, agentic architectures, and reinforcement learning, bringing new ideas directly into production.
What They’re Looking For
Production AI Builders
3+ years building and shipping real AI systems, ideally agent-based or complex ML systems in production.
Strong ML Engineering
Hands-on experience with Python, PyTorch, and modern ML workflows.
Post-Training / RL Experience
Experience with RLHF, DPO, reward modeling, or similar post-training techniques is a strong plus.
Startup or Founder DNA
Former founders, founding engineers, or builders who have shipped technically complex products.
Technical Depth
Strong background in computer science, computer engineering, or electrical engineering (top university backgrounds preferred).
Bonus Experience
Nice-to-have but not required:
-
Silicon verification workflows
-
EDA tooling (Synopsys, Cadence, Questa)
-
Publications in top conferences (NeurIPS, ICLR, DAC)
Why This Role Is Interesting
Hard Problems, Real Impact
Your work will reduce multi-day chip debugging workflows into minutes using AI agents.
Rare AI + Hardware Intersection
Work at the frontier of agentic AI applied to one of the most complex engineering domains in the world.
True Founding Role
Own the core intelligence layer of the product and help shape the direction of the company.
Production from Day One
Your systems will be used by real semiconductor teams solving real problems.
Competitive Compensation
$175k – $250k base + meaningful founding equity (0.4–0.75%).
Ideal Background
You might be a strong fit if you have:
-
Built agentic AI systems from scratch
-
Designed RLHF / evaluation pipelines
-
Shipped technically complex production AI systems
-
Previously worked at a startup, research lab, or as a founder
-
Enjoy building messy, ambitious systems from zero to one
Location
San Francisco (Onsite)
H1B transfers supported for strong candidates.
If you want to work on frontier AI systems solving real engineering problems, this is one of the most technically interesting roles in the market right now.
| Location: | San Francisco, CA |
|---|---|
| Job type: | Permanent |
| Emp type: | Full-time |
| Salary type: | Annual |
| Salary: | USD $225,000.00 |
| Job ID: | 35556 |