A local RAG system with an integrated artificial intelligence (AI) engine that turns papers, reports, tables, figures, audio and video into a searchable scientific knowledge base. It runs on the user's machine to protect data privacy, with no paid API required for normal use and no need to send documents to external services.
EcoLab RAG lets a group load documents, verify index status, search in natural language and explore thematic connections across the corpus.
Video recorded from a local EcoLab RAG installation. The demo shows the dashboard, supported upload formats, semantic search, result ranking and opportunity detection in the thematic graph.
EcoLab RAG is built for groups that already produce scientific content but need to find evidence, compare results and reuse knowledge without endless manual searches.
Query the corpus in natural language and retrieve relevant fragments with ranking, similarity and bibliographic metadata when available.
The system extracts tables, figures and images so they can be located and reviewed without opening every document manually.
It transcribes audio and video locally, indexes the content and lets the user return to the relevant minute of a result.
Builds concept maps, document communities and research opportunities from relationships inside the corpus.
Includes an integrated AI engine that runs locally with PostgreSQL and local models. Normal use does not require sending documents to external services or paying APIs.
Delivery includes installation, documentation and training so the group can load, search, delete and maintain its own corpus.
EcoLab RAG is not delivered as a SaaS account. It is installed, configured and transferred to the group as a reproducible workflow.
We review corpus type, formats, available machine, confidentiality constraints and intended use.
Docker, PostgreSQL/pgvector, local models, launcher and web interface are configured.
We load a test set and validate search, tables, figures and multimedia content.
A practical session for uploading documents, interpreting results, reviewing logs and maintaining the index.
Post-delivery support for installation issues, initial usage and workflow adjustment.
Final pricing depends on environment, number of machines, initial corpus, required support and customization level. These ranges help align expectations before a demo.
For an individual researcher, PI or small lab that needs one functional local installation.
For groups with several users, a real corpus and need for team handover.
For centers, units or labs that need a customized deployment and support plan.
There is no mandatory monthly fee to use the system. Extended support, new customizations or additional deployments are quoted separately.
EcoLab RAG is a local product. That improves privacy and control, but it requires a suitable machine and an initial installation process.
Runs on macOS and Windows with Docker Desktop. Performance depends on CPU, memory and corpus size.
Once local models are downloaded and configured, daily use can run offline for local documents.
Audio and video are processed locally. A long video can take several minutes, and heavy files should be uploaded one at a time.
In a demo we review document types, corpus size, installation requirements and whether the right fit is Lab, Research Group or Institutional.