We build tools where the math is the model — a local-first coding agent that verifies its own work, and a zero-shot engine that screens molecules from topology alone.
Everything we ship is downstream of these. They are not marketing — they are the filter on every design decision.
We start from equations, not datasets. Where others train on the past, we derive from structure — topology, geometry, information theory. The math is the model.
Software that runs on your machine, against your keys, on your terms. No silent telemetry, no rented intelligence. Your work never leaves your custody unless you choose.
A small lab shipping tools we use ourselves. No engagement metrics, no growth loops. We measure success in decades of useful software, not quarters of attention.
Different domains, identical thesis: trust the math, run locally, verify against ground truth.
A local-first autonomous coding agent that plans, writes, and verifies code across multiple models. Its hostile verifier launches a real browser and catches what other agents ship broken.
A zero-shot topological alignment engine for thermodynamic inference. Pure mathematics — no training data. The SPANDA engine screens molecular geometry for medicine, batteries, and materials.
Our two products are surfaces of a deeper research program. Each asks the same question from a different angle: can structure replace examples?
In code, the structure is the runtime — verify against it. In molecules, the structure is the topology — align against it. Neither requires a training set. Both require the math to be right.
Persistent-homology descriptors that let a model reason about molecular function from geometry alone — no labelled examples, no fine-tuning.
Treating code output as adversarial: launch real runtimes, probe endpoints, fail loudly. Verification is not a second model — it is ground truth.
Agent loops that plan, execute, and recover on-device. Long-horizon memory and crash recovery that resumes exactly where the queue stopped.
We take on a small number of research and engineering engagements where the problem is hard and the math matters. Reach out.
lab@bhaskara.io