Labs enter through a controlled workspace
Team members accept invites, authenticate with Supabase-backed email auth, and land in a workspace built for their lab instead of a generic dashboard.
AI-native operating layer for modern labs
Stop stripping away experiment context before AI ever sees it.
Labframe connects instruments, artifacts, notes, parameters, and run lineage into one operating surface. Researchers can upload data or attach a live run, keep the full setup intact, and get analysis or troubleshooting without re-explaining the experiment every time.
2seeded experiment tracks
3legacy runs imported
2connector surfaces
What researchers see
One product, three operational loops
- Analyze new uploads and live runs with the full context ledger visible.
- Replay every trial in the Run Trail with lineage, notes, and prior analysis intact.
- Guide each workspace through AI onboarding so the system knows the lab’s tools and goals.
Concierge setup turns lab context into product structure
The onboarding agent captures instruments, experiment families, failure modes, desired outputs, and collaboration needs, then turns them into templates, connectors, and next actions.
Every run becomes a reusable context package
Data artifacts, notes, parameters, results, and analysis summaries persist into a shared timeline so later runs have continuity instead of amnesia.
Where data becomes an AI-native workspace
Left rail for intake, center canvas for analysis, right rail for “context in scope” so users can see what the AI is using.
Where labs replay every decision
Timeline plus table filters for experiment, status, owner, tags, instrument, and date, with direct links into run detail and comparisons.
Where ingestion becomes operational
Live telemetry and upload-based flows coexist so labs can start with files and graduate to connected instruments without replatforming.
Auth, Postgres, Storage, Realtime
Invite-only access, structured experiment records, private artifact buckets, and workspace-scoped row-level security are built into the implementation.
Web app on Vercel, Python for instrument work
The new Next.js product layer handles auth, UX, and product APIs while the existing Python stack keeps device communication and ingestion responsibilities.
Scaffolded for Vercel + Railway/Render
This repo now includes a deployable `web/` app, Supabase SQL migrations, and import / internal ingest paths for the legacy workbench data.
Request access
Need an invite for your lab?
Use the invite flow if your workspace already exists, or contact founders@labframe.ai to seed a pilot workspace and migrate your first experiment history.