Workspace intelligence

Chat with your data. Invite your team. Ship faster.

Upload documents, index into a private vector store, and run a grounded copilot per workspace — with IAM, email invites, and full control over LLM keys.

Free to self-host — you bring Postgres, S3, and optional Resend for invite emails.

From files to answers
Presigned uploads, text extraction, chunking, and OpenAI embeddings — then similarity search at query time so the assistant cites your real numbers and tables.

Ingest

CSV, JSON, Markdown, plain text — same pipeline from chat or documents.

Collaborate

Workspaces, roles, and email invites with accept links.

Workspace-native RAG
Documents in S3, chunks in pgvector, retrieval always scoped by workspace — no cross-tenant leakage.
Models you control
OpenAI, Anthropic, Gemini, and more — multiple keys per workspace, one default for chat.
IAM & invites
Invite teammates by email, accept with a secure link, and manage roles from a dedicated IAM page.

Everything in one stack

Next.js 16 · PostgreSQL · pgvector · S3 · Auth.js · Prisma

Charts & insights
Structured answers with optional charts, plus background insight scans on your corpus.
Security-first
Presigned S3 access, RBAC, audit-friendly patterns, and encrypted API keys at rest.
Insight engine
Heuristic and extensible scans for anomalies, risk language, and sentiment shifts.

Ready to create a workspace?

Sign up, name your first workspace, then invite your team from IAM.