Skip to content

Neuronautix · Products

R&D-derived products for preclinical, biotech, and scientific teams

Neuronautix is not only a consultancy. Our open-source research and ontology work feeds three productized offerings — built once, deployed many times. Each one packages a piece of our R&D into something a team can buy, install, or subscribe to without a custom engagement.

Product · Local-RAG knowledge stack

FounderGraph-Lab: private LLM + RAG over your company's knowledge

A turnkey local deployment of a multi-agent LLM stack with retrieval over your internal corpus — SOPs, study reports, regulatory documents, code, slide decks, lab notebooks, partner contracts. Runs on-prem or in a private cloud. No data leaves your perimeter.

Built on the same multi-agent architecture as Neuronautix's Science Agent Squad, with an ontology-aware retrieval layer (HCMO, MBO, NAMO) that makes domain queries far more precise than naïve vector search.

  • Local LLM hosting (Ollama, vLLM, or your preferred runtime)
  • RAG indexing of documents, code, structured data, and meeting notes
  • Ontology-constrained retrieval and query rewriting for scientific corpora
  • Multi-agent task pipelines (literature review, regulatory drafting, evidence map)
  • Audit log + provenance: every answer cites the source chunk
  • Maintenance, model updates, and quarterly review subscription

Product · FAIR Virtual Control Group reports

FAIR-VCG: replace control animals with auditable historical evidence

Send us a historical control dataset (CSV, study reports, telemetry exports). We return a FAIR-compliant Virtual Control Group report: matched historical controls, statistical comparability checks, provenance trail, and the metadata package needed to defend the choice in a regulatory submission.

Built on the open-source FAIR-VCG pipeline, with LLM-assisted curation and ontology-anchored metadata to keep every claim traceable to a source record. Aligned with 3Rs animal-replacement principles and emerging NAM / IND data expectations.

  • Per-study report: matched controls, statistical justification, FAIR metadata bundle
  • Provenance graph linking every value to its source dataset
  • Auditable LLM curation log (model, prompt, retrieval set, human review)
  • Optional follow-up: integration into your study workflow as a recurring service

Product · Ontology support contracts

HCMO & MBO: licensed ontology extensions and validation for vendors

The Home-Cage Monitoring Ontology (HCMO) and Mouse Behavior Ontology (MBO) are open-source semantic vocabularies maintained by Neuronautix and collaborators. For technology vendors, platform owners, and consortia, we offer paid support contracts on top of the open base.

  • Custom vendor-specific ontology extensions (your sensors, your readouts, your schema)
  • SHACL / ShEx validation profiles for your data export format
  • Validation API: paste a metadata file, get a conformance report
  • Ontology-aware LLM constraint layer (for AI agents navigating your data)
  • Roadmap input, prioritized issues, and a Neuronautix point of contact

Why these three

Each product is a productized slice of our R&D

Neuronautix runs active R&D in three areas — multi-agent scientific reasoning, ontology-constrained LLM navigation (with LIRMM, Dr. Konstantin Todorov), and neuro-AI / stat-AI embedding methods for preclinical knowledge generation. The products above package the most mature, deployable outputs of that research. Buying a product also funds the next iteration of the open-source stack.

Start with a scoping call

Which product fits your team?

Tell us about your corpus, your study, or your platform — we'll come back with a fit assessment, an indicative scope, and a quote.

  • Email: neuronautix [at] gmail.com
  • Phone: +33 (0)7 85 30 64 70
  • Montpellier, France