The Wawe story: from two agents to one platform

Wawe AI didn't start as a platform. It started as a single agent that could answer one hard question: given a protein sequence, what can we say about it?Three years later, that agent is one of four Spaces in a unified drug discovery environment. Here's how we got here.

2023 — The founding bet

We founded Wawe on a single conviction: research-grade computational drug discovery shouldn't require a PhD in bioinformatics to operate. The tools existed — ESMFold, AlphaFold, AutoDock, GROMACS — but each demanded its own file formats, flags, and folklore. The bottleneck wasn't the science. It was the interface.

2024 — The Biology Agent

Our first release wrapped 100+ computational biology tools behind a natural-language interface. Ask about a target, a fold, or a pathway in plain English; the agent picked the tools, ran them, and explained the result. It was the first time a wet-lab biologist could run a structure prediction without touching a command line.

2024 — The Chemistry Agent

Biology answers "what should we target?" Chemistry answers "what should we make?" So we built ChemGraph — ADMET prediction, retrosynthesis planning, and molecular docking, plus the explainable SAR analysis that tells you why a molecule is potent. Two agents now covered both halves of early discovery.

Two strong agents that couldn't talk to each other is just two silos. The real product was the seam between them.

2025 — One platform

We merged Bio and Chem into a single LangGraph-based platform with streaming results, session memory, and full tool tracing. A target discovered in Biology now flows directly into a design brief in Chemistry — no copy-paste, no context lost. The pipeline became continuous.

2026 — Enterprise pilot

This year we deployed to Azure with enterprise infrastructure, authentication, and role-based access for research institutions. The Work Station added precision docking and MD; Wawe Grid added the operations layer to track every run across labs. What began as one agent is now an end-to-end environment spanning 190+ tools.

Where we go next

The through-line has never changed: collapse the distance between a research question and a grounded answer. Shared project spaces, deeper lab integration, and tighter feedback between design and validation are next. If you're a researcher who wants in, request access — the pilot is open.