Startup Ideas Bank

Open-source AI workbench with shaky real-world adoption prospects

AI roast score: 65/100 (C)

The idea

An ambitious AI workbench for scientific research that promises the moon but might not get scientists to even look up.

In the founder's own words

synthetic-sciences/openscience — The open-source AI workbench for scientific research

The open-source AI workbench for scientific research

Give it a goal. It reads the literature, writes and runs code, runs the experiments, and writes up what it found.





Install · Quickstart · Docs · Atlas


OpenScience is an AI workbench for scientific research. You give it a goal, and it works through the research loop the way a capable collaborator would. It reads the papers that matter, forms a hypothesis, writes and runs code, runs experiments on real compute, queries the major scientific databases, and writes up the result. It runs as a workspace in your browser and works with any frontier or open-weight model from Anthropic, OpenAI, Google, and dozens of other providers, using your own API keys. No account is required.

It is model-agnostic, open source, and built to do real work in machine learning, biology, physics, and chemistry.

What it does


Runs the whole loop. Literature review, hypothesis, code, experiment, analysis, and write-up, in one continuous session.

Research agents. A research agent by default, plus biology , physics , and ml specialists, with critique and literature-review sub-agents and a read-only plan mode.

290+ skills. Training (DeepSpeed, PEFT, TRL), evaluation, dataset work, molecular and clinical biology, cheminformatics, papers and LaTeX, figures, and cloud compute (Modal, Tinker, and others).

Scientific databases as tools. UniProt, PDB, Ensembl, ChEMBL, PubChem, arXiv, OpenAlex, Semantic Scholar, and around 30 more, queryable directly by the agent.

A real workspace. A browser UI with a file tree, an editor, a terminal, session history, and inline rendering for molecules, structures, genomes, and plots.

Extensible. LSP integration, MCP servers, plugins, custom agents and commands, and a TypeScript SDK.

Install

Install with npm, then open the workspace:

npm install -g @synsci/openscience
openscience

The command is openscience , and it opens the workspace in your browser. The first time you run it, a short setup walks you through how to power the models — Atlas managed models, your own provider keys, or skip and start on the free demo models. If you would rather not install it globally, npx synsci does the same thing in a single step:

npx synsci

Platform binaries are also attached to GitHub Releases ; see the changelog for what's new in each version.

Quickstart

Set an API key from any provider ( ANTHROPIC_API_KEY , OPENAI_API_KEY , GEMINI_API_KEY , and so on) and start the workspace:

export ANTHROPIC_API_KEY=sk-ant-...
openscience

openscience opens the workspace in your browser. Your keys stay on your machine and requests go straight to the provider. You can also run openscience keys add

The roast

Your open-source AI workbench sounds like a researcher's dream, but dreams and reality rarely coincide. Integration with 'frontier or open-weight models' and '290+ skills' seems over-engineered for a market that resists change due to stringent validation needs and conservative adoption rates. Your solo-founder status (q13=solo) and no funding (q14=no_funding) further jeopardize your chances of delivering and maintaining such a complex platform. Your reliance on users to supply their own API keys makes setup a hassle, especially for a target audience of 'professionals' (q5=professionals) who are often not tech-savvy. Moreover, your 'product-led' growth strategy (q17=product_led) clashes with the need to validate the core hypothesis that researchers 'will pay' (q15=will_pay) for a tool they traditionally build in-house. You list an impressive array of features, but 'no account is required' and 'runs in your browser' are not enough to overcome the inertia in scientific research workflows. Scientists are notoriously skeptical, and your solution might be too broad and unfocused to gain serious traction.

Red flags

Verdict

Your AI workbench needs a reality check on adoption barriers and user onboarding pain points.

Roast your own startup idea →

More ideas roasted by AI

Browse all startup ideas →