A small lab with a long-term view
We are researchers, engineers, and builders who believe the companies closest to a problem should own the AI that solves it. This is who we are and how we work.
Frontier AI should belong to the people who use it.
We started Alamo Research because the most capable AI in the world was concentrating in a handful of labs, while the companies with the hardest, most valuable problems were being handed black boxes and rate limits. We believe the opposite future is possible: one where the teams closest to a problem own the models that solve it.
So we hold ourselves to three commitments. We do real research, not integration work dressed up as research. We measure everything, because a model that cannot be evaluated cannot be trusted. And we hand over the keys — weights, data, and pipeline — because a partnership that creates dependency is not a partnership.
This is slower and harder than selling an API. It is also the only way we know to build AI that lasts longer than a hype cycle.
The people behind Alamo Research.
Adeen Shukla
Co-founder & CEO
Sets the direction of the lab and keeps every engagement anchored to one question: does this move the frontier for our partners?
Deepak Joshi
Co-founder & CTO
Leads the technical organization, from training infrastructure to the deployment pipelines that carry models into production.
Suyash Dixit
Chief AI Officer
Owns the research agenda — training recipes, evaluation harnesses, and the agents that run on the models we build.
Lavleen Gonzales
HR Manager
Builds the team behind the lab, making sure the people doing frontier work have the environment to do it well.
Vinayak V
Technical Lead
Turns research into systems that hold up under real-world load, leading engineering across our client engagements.
Mansi Joshi
Chief Design Officer
Shapes how our work looks and feels — from the interfaces our models live in to the way we communicate research.
How a small lab does frontier work.
Small teams
Every engagement is run by a small, senior team with full context — no hand-offs, no telephone game.
Write it down
Decisions, experiments, and results live in documents anyone can read. Institutional memory beats hallway memory.
Ship weekly
We work in short loops with visible output, so progress is something you can inspect rather than take on faith.
Disagree openly
The best idea wins regardless of title. We argue in the open, commit fully, and revisit when the data says to.
The motivations behind the work.
Frontier, applied
We are motivated by the gap between what models can do in papers and what they do in production. Closing it is the job.
Evidence over opinion
Rigorous, task-specific evaluation is how we know we are right. We would rather be measured than impressive.
Partners, not vendors
We embed with the teams we work with. Their domain expertise plus our research is the whole model.
Ownership matters
AI that companies own — weights, data, and pipeline — is AI they can trust, extend, and build a future on.
Let's build your AI at the frontier.
Partner with Alamo Research to train, fine-tune, and deploy models built for your domain.