Drug discovery takes too long and costs too much

The pharmaceutical industry has not meaningfully improved its core R&D process in decades. The result is a pipeline that is slow, expensive, and increasingly unreliable.

12+ yrs

Average time from molecule discovery to market approval

$2.6B

Average cost to develop a single new drug, including failed candidates

90%

Of drug candidates that enter clinical trials never reach patients

AI-accelerated molecular screening

Helix uses deep learning models trained on 40 million molecular interactions to predict drug-target binding, toxicity, and efficacy before a single wet-lab experiment runs.

A market expanding faster than R&D budgets can keep up

One platform, from target to candidate

Input

Protein structure upload

Accepts PDB files, AlphaFold predictions, or FASTA sequences. Auto-detects binding sites.

Engine

Multi-model screening

Runs binding, toxicity, and ADMET prediction in parallel. Returns ranked candidates in under 4 minutes.

Output

Lab-ready candidate list

Exports to CSV or direct integration with CRO ordering systems. Includes synthesis cost estimates.

Revenue is growing 18% month over month

Six paying customers across pharma and biotech, with two multi-year contracts signed in Q2.

Three revenue streams, one platform

Builders who have done this before

Dr. Anya Krishnan

Co-founder & CEO

Former principal scientist at Genentech. PhD in computational biology from Stanford. 14 patents in drug-target prediction.

Dr. Marcus Lindqvist

Co-founder & CTO

Built ML infrastructure at DeepMind for 6 years. Led the protein folding pipeline team for AlphaFold 2.

Dr. Felipe Nakamura

Chief Scientific Officer

20 years at Novartis leading small molecule discovery. Brought 3 drugs from discovery to Phase II trials.