Member of Technical Staff - Research Scientist
Architect Labs
- Palo Alto, CA
- On-site
- Senior level
- 3+ yrs experience
- ML Research / Applied AI Research
- Visa: not specified
The role in 30 seconds
Architect Labs is a frontier AI lab for custom silicon, partnering with frontier labs, clouds, and advanced fabs to tape-out custom chips co-designed for next-generation AI workloads. As a Member of Technical Staff, you sit at the intersection of chip design and frontier AI, translating deep hardware engineering expertise into agentic systems that can reason about, generate, and verify real silicon. You will research self-improving agent harnesses, build evals and environments for chip-design tasks, and partner with ML research and infra teams to turn hardware domain expertise into reward signals and production systems.
Is this role a fit for you?
The backgrounds this role is most likely to hire from, as RoleStack reads the posting.
Strong match
- Data platform / ML pipeline
- B2B infrastructure / API / developer tools
- Startup / greenfield
- Growth-stage
Workable stretch
- B2B enterprise platform / workflow software
- Creator / builder / self-serve tool
- Enterprise-scale
Harder transition
- B2C consumer-facing application
- Internal tooling / back-office systems
- Regulated environment
Industry: Semiconductors / AI hardware (chip design). See your personal match score
What they're really screening for
Must-haves
- CORE PhD in Computer Science or a related field, or 3+ years of research experience at a frontier AI lab
- CORE Experience designing and improving agent harnesses, including orchestration, tool use, memory, feedback loops, and recovery mechanisms for autonomous long-horizon tasks
- CORE Research instinct: ability to stay current with frontier AI research, critically assess new approaches, and turn promising findings into measurable improvements
- CORE Exceptional systems engineering: architect and ship reliable, large-scale infrastructure that others build on, holding research code to production standards
- PLUS Builder mindset: own ambiguous problems end to end, prototype quickly, turn promising ideas into production systems; pragmatic, resourceful, comfortable moving fast
- PLUS Interest in hardware: excitement about applying frontier AI to chip design and willingness to learn deeply from hardware engineers
- CORE Experience applying frontier AI to chip design / silicon engineering domain or adjacent hardware engineering domain
Nice-to-haves
- PLUS Research published at leading top conferences (ICLR, ICML, NeurIPS, etc.)
- PLUS Meaningful open-source contributions to coding agents, agent harnesses, or ML tooling, or building a harness from scratch
- PLUS Designing and deploying production multi-agent workflows and agents that run autonomously for days or weeks
- PLUS Building evaluations and benchmarks for complex engineering tasks
- PLUS Familiarity with developing reward signals, post-training, or reinforcement learning methods for language models
- PLUS Familiarity with training or serving transformer-based models and building high-performance, large-scale ML systems
Résumé keywords for this role
Terms the posting stresses. Include the ones you can honestly back up.
Tools & technologies
- ICLR
- ICML
- NeurIPS
- post-training
- reinforcement learning
- transformer-based models
- ML systems
- agent harnesses
- multi-agent workflows
- evals
- benchmarks
- reward signals
- tool use
- memory
- feedback loops
Concepts & phrases
- Member of Technical Staff
- Research Scientist
- frontier AI lab
- custom silicon
- agent harnesses
- self-improving harnesses
- swarm of agents
- long-horizon tasks
- reward signals
- training signals
- chip-design tasks
- 0 to 1 environment
- production systems
- large-scale infrastructure
- research code to production standards
Words that count as the same thing: "agent harnesses" ≈ agent frameworks, agent orchestration systems, agent scaffolding · "frontier AI lab" ≈ leading AI research lab, top-tier AI company, cutting-edge AI research organization · "post-training" ≈ fine-tuning, RLHF, model alignment training · "reward signals" ≈ reward modeling, RL reward design, preference signals · "evals" ≈ evaluations, benchmarks, model assessment
Check my résumé against these keywordsHiring at Architect Labs
Open senior roles RoleStack tracks right now, posted in the last 30 days.
Full job description from Architect Labs
From Architect Labs's posting. View the original
ABOUT ARCHITECT LABS
Architect is a frontier AI lab for custom silicon. We partner with frontier labs, clouds / neoclouds, physical AI companies, and advanced fabs to tape-out custom chips co-designed for next-generation AI workloads. Our goal is to compress end-to-end software to silicon timelines, and maximize intelligence per watt and per dollar for the world. We are a small exceptional team across silicon, systems, software and frontier AI. Our team have led research teams at nearly every frontier AI lab, and at some of the most complex SoCs in the world.
Show full description Show less
What You'll Do
As a Member of the Technical Staff at Architect Labs, you'll sit at the intersection of chip design and frontier AI translating deep hardware engineering expertise into agentic systems that can reason about, generate, and verify real silicon.
- Research on self-improving harnesses to guide a swarm of agents work autonomously for days / weeks. Stay on top of SOTA frontier research.
- Partner closely with the ML research, post-training, and infra teams to turn hardware domain expertise into reward signals, benchmarks, and training signals.
- Build scalable and robust software stacks that transform ideas into production systems seamlessly.
- Build evals and environments that tackle chip-design tasks, grounding model behavior in real hardware engineering work.
- Move fast in a 0 to 1 environment: prototype, dogfood, break things, iterate. Translate ambiguous chip-design challenges into concrete agent capabilities that ship.
What We'd Like to See
Qualifications & Skills:
- Background: PhD in Computer Science or a related field, or 3+ years of research experience at a frontier AI lab.
- Agent Harness Research: Experience designing and improving agent harnesses, including orchestration, tool use, memory, feedback loops, and recovery mechanisms that enable agents to work autonomously on complex, long-horizon tasks.
- Research Instinct: Ability to stay current with frontier AI research, critically assess new approaches, and turn promising findings into measurable improvements.
- Software Engineering: Exceptional systems engineering. You architect and ship reliable, large-scale infrastructure that others build on, and you hold research code to production standards.
- Builder Mindset: Ability to own ambiguous problems end to end, prototype quickly, and turn promising ideas into production systems. Pragmatic, resourceful, and comfortable moving fast.
- Interest in Hardware: Excitement about applying frontier AI to chip design and a willingness to learn deeply from hardware engineers.
Strong candidates may also have experience with:
- Research published at leading top conferences (ICLR, ICML, NeurIPS, etc).
- Meaningful open-source contributions to coding agents, agent harnesses, or ML tooling, or building a harness from scratch.
- Designing and deploying production multi-agent workflows and agents that run autonomously for days or weeks.
- Building evaluations and benchmarks for complex engineering tasks.
- Familiar with development reward signals, post-training or reinforcement learning methods for language models.
- Familiar with training or serving transformer-based models and building high-performance, large-scale ML systems.
What We Offer
- Competitive salary and meaningful equity stake
- Fast-paced startup with autonomy and visible impact
- Cutting-edge AI-driven chip design challenges
FAQ
How much experience do you need?
The posting asks for 3+ years in research at a frontier ai lab. It is a senior-level role.
Does Architect Labs sponsor visas for this role?
The posting does not say whether visa sponsorship is available. Ask the recruiter before you apply.
Is this role remote?
It is an on-site role.
What tools and technologies does the posting mention?
ICLR, ICML, NeurIPS, post-training, reinforcement learning, transformer-based models, ML systems, agent harnesses, multi-agent workflows, evals, benchmarks, reward signals, tool use, memory, feedback loops.