Transform documents into searchable knowledge with AI-powered cited answers
Agent wiki and visual workspace for Claude Code, Codex and other AI coding agents
Stage
Established
Not given
Founded
2003
Not given
Based in
US
Not given
Pricing model
Freemium
Freemium
Pricing
SaaS: $30 per user per month (10% discount billed annually), includes 5GB storage, 30+ agents, SSO and audit logs. 15-day free trial with no credit card required. On-Prem Edition pricing available by contacting sales.
Free: Nimbalyst app and personal wiki. Wiki Teams: $20 per user per month or $200 per user per year (currently free during beta).
Key features
AI-powered document Q&A with page/row/paragraph references
30+ pre-built AI Agents for contracts, compliance, invoices and more
Multi-source document processing (PDF, DOCX, XLSX, PPTX, CSV and more)
Connect Google Drive and Microsoft SharePoint with auto-sync
Query SQL databases in plain English without SQL skills
Website content indexing for live knowledge sources
Bring your own LLM (OpenAI, Anthropic, Gemini, Llama)
Role-based access control (RBAC) and full audit logs
On-premises or private cloud deployment options
Custom Agent creation on user-defined templates
WYSIWYG markdown editor with AI-assisted editing
Red/green inline diffs for AI changes approval
Visual editors for mockups, diagrams, and drawings
Parallel agent session management
File and session linking
Persistent searchable session history
Pull request review in Developer Mode
MCP server support
What makes it different
Every answer is cited to its source document—no hallucinations or invented claims
Delivers finished work products instead of chat transcripts
Data stays on your infrastructure with full deployment control
Built for regulated industries with permission-aware design and full audit trails
Choose and change your LLM independently from the platform
Agents write the wiki as they work, eliminating manual documentation
Open source and local-first: your data stays on your machine
Built by Evergage founders who scaled to $250M ARR
One unified workspace instead of multiple disconnected tools