Think Different. Build the Future. đ
Our Mission
Build everyday AGIâtrustworthy, consumer-grade agents that redefine humanâAI collaboration for millions. Software shouldnât wait for commands; it should partner with you, amplifying what you can do every single day.
Why AGI, Inc.
Weâre a stealth team of elite ex-founders, Stanford and Berkeley researchers, and DeepMind alumni. With years of agent research, our platform already runs live agent workflows with 99.9% reliability. Our investors funded the first wave of AI giants; now theyâre backing us to deliver the nextâ
everyday AGI.
If you see possibility where others see limits, read on.
About The Role
Youâll be among our very first researchers, turning bold hypotheses into production-grade capabilities and shaping the scientific backbone of our core product.
Expect to:
- Push frontiers. Prototype new architectures in reasoning, long-horizon planning, memory, multi-agent coordination, and alignment.
- Ship, donât shelve. Run live A/Bs on billions of real trajectories and see your ideas land in production within weeks.
- Bridge theory and delight. Pair with product engineers to fuse novel algorithms with delightful consumer experiences.
- Define the bar. Establish our research rigor, safety metrics, and evaluation pipelines for everything that calls itself âagentic.â
- Mentor and magnify. Level up interns, residents, and mixed-discipline hackers, multiplying the impact of the entire team.
Key Responsibilities
- Create novel methods. RLHF, scalable supervision, tool-use planning, state-space memoryâwhatever unlocks new agent breakthroughs.
- Scale experiments fast. Design datasets, train across thousands of GPUs, analyze results, iterate.
- Own evals. Build automatic, adversarial, and human-in-the-loop benchmarks that predict real-world reliability.
- Publish with purpose. Contribute to NeurIPS/ICML/ICLR/ACL when it moves the fieldâand our productâforward.
- Harden for safety. Develop guardrails and interpretability tools so users can trust what the agent does next.
You Bring
- MS or PhDâor equivalent track recordâin ML / NLP / RL / systems, with 3+ strong papers or open-source impact.
- Mastery of PyTorch or JAX and large-scale training across distributed GPU clusters.
- Hacker mindsetâequally happy debugging HTML selectors and re-deriving tensor math.
- Passion for shipping to real users, not just writing PDFs.
- Love for in-person collaboration; ready to build in SF.
Our Culture
đ˘
All in, in person. Sparks fly faster face-to-face.
đ
Ship by default. Novel and polished can coexistâspeed is the feature.
đ¤
One band, one sound. Radical candor, zero politics, help each other win.
Our perks
đ˝ Lunch, dinner and snacks at the office.
đĽ Fully covered medical, dental, and vision insurance for employees.
đŚ 401(k).
âď¸ Relocation and immigration support.
How to Apply
Send us:
- A linkâor 60-sec videoâof your most meaningful paper, repo, or demo and why it matters.
- Your resume or Google Scholar / LinkedIn.
- Two sentences on the toughest research problem youâve cracked.
Every exceptional candidate hears back within
48 hours.
We canât wait to see what youâve gotâand to welcome you to the AGI Inc. team!
