Today, we introduce Rafflesia. We are building the infrastructure for life science research agents.

It began with a simple question: given the right tools, can an agent reproduce the conclusions of a research paper from its raw data? The answer was no: the tools the agent needed didn't exist.

Agents are extraordinarily capable, but they inherit a world that was never built for them. Performing sequence homology, downloading a dataset, running a molecular dynamics simulation, visualizing a result. Every task becomes a meta-quest of its own. Rafflesia is built to accelerate the hypothesis-to-theory-to-experiment loop that creates knowledge.

The perfect toolbox will not be a stitching-together of the services humans already use. It will be new, and built for agents from the ground up. Headless interfaces. Retrieval systems that answer without bloating the context window. Curated databases that live in the cloud and can be queried blazingly fast. Wet lab can be part of these tools.

Our iteration loop is simple:

Structural biology studies the 3D structure, interactions, and dynamics of proteins and the complexes they form. This foundational understanding drives the design and discovery of better drugs.

We're tackling this problem because it has some unique characteristics.

First, the field produces experimental results at an exponential scale. Storage and processing systems were designed for local HPC centers tied to research labs. The infrastructure to index, curate, organize, and serve this data needs to be rebuilt for the cloud and AI agents.

Second, structural biology is one of the most verifiable environments in all of science. The feedback loops are everywhere you look: stoichiometry, hydrophobic and ionic forces, primary, secondary, and tertiary structure, bond geometry, angstrom-level positions you can measure directly. These systems are also easy to simulate; the laws of physics provide the RL environment.

The problems also span the full range of difficulty. Some are simple enough for an agent to handle today. Others sit at the edge of what anyone understands, like large multi-protein complexes. That diversity makes the field ideal for agentic hill-climbing, where each solved problem opens the next.

The payoff is enormous. If we can make AI genuinely good at protein engineering and nanoscale design, the ceiling is hard to see. Imagine an AI that could design something like the nuclear pore complex or a ribosome.

Open. Our goal is to move biology out of the walled garden. We contribute to open source wherever we can.

Ship progressively. We believe moonshots are reached part by part, iteratively, not through inspirational, fuzzily-defined goals. Build the piece, ship the piece, then build the next one. We also believe that’s the way safe AI systems for biology are built.

Software first. The data already exists, and excellent wet labs already generate more of it. Our job is the software layer around that decentralized network: standardized procedures, reproducible datasets, and tight loops that let better algorithms learn from real evidence.

Cloud native. Building for the cloud is building for AIs. It is also building for scale. Much of today’s scientific computing was designed for HPC. We think there is a huge opportunity to reinvent it for the cloud era.

We have already launched:

• Rafflesia Homology, a serverless search engine built on object storage for homology search, as precise as Foldseek and MMseqs but much cheaper and more scalable.
• Headless Mol* and headless MDsrv, command-line tools that let agents create molecular visualizations without a human in the loop.
• TinkerBench, a benchmark for measuring how well agents perform on real structural biology tasks.

We are assembling a world-class team of engineers and scientists to build groundbreaking infrastructure for life science. Join us.