Native SQL and query planning
The platform is building SQL over governed company objects, with parameters, explain plans, and policy-aware reads. In the hosted public beta, start with cited questions over Files and Obsidian.
Pantasso Data Platform
Current operating context for your company and its AI.
This is a hosted public beta, not a finished platform. Start with Files or Obsidian, try the synthetic demo, then connect Slack, Drive, Telegram, or GitHub with founder-led setup. Cited answers come with receipts back to the approved record.
How the platform works
From one approved source to trusted answers and action.
Pantasso follows a five-stage path. Each stage is visible, auditable, and under the workspace owner's control.
1
Connect approved sources
The hosted public beta starts with approved files. Slack, Drive, Telegram, and GitHub can be connected. Other messaging, Google Workspace, Microsoft 365, email, calendars, spreadsheets, admin directories, code and project systems, CRM, finance, commerce, support, observability, cloud audit, files, databases, webhooks, SDKs, and structured imports families are listed for later setup, not treated as a live suite.
2
Maintain operating context
Pantasso organizes approved activity into governed company records, evidence, decisions, and work while preserving the original source, authority, author, and time.
3
Ask questions with evidence
People and assistants can ask natural-language questions and receive cited answers. Teams can also use structured queries and views when they need deeper analysis.
4
Give people and assistants clear boundaries
Choose what each person or assistant may read. Configured governed workflows also define what an assistant may propose or change, with recorded approval and result evidence.
5
Keep the company state healthy
See source freshness, failed syncs, activity, and backups so company context stays current and useful.
Built for real company work
Keep the work itself connected to evidence and action.
Pantasso maintains the knowledge, structure, controls, history, and agent access your team uses to understand company state and move work forward.
Files first, then connected sources, then listed families.
The live first source is Files and Obsidian. Slack, Drive, Telegram, and GitHub need a connection. The directory lists 36 families, including Slack, Microsoft 365, Google Workspace, Gmail, Calendar, Sheets, Admin Directory, Telegram, GitHub, GitLab, Notion, Linear, Jira Cloud, Salesforce, HubSpot, Stripe, Shopify, QuickBooks Online, Xero, Zendesk, Intercom, PagerDuty, Datadog, Sentry, Cloudflare, AWS CloudTrail, Vercel, Files and Obsidian, Supabase, Structured Database, Signed Webhooks, Connector SDK, and MCP-compatible JSON imports. A listing is not a live or GA connector.
Source ingestion
- Live first source: Files and Obsidian
- Needs connection: Slack, Drive, Telegram, GitHub
- Other families stay listed until they are proven live
- Each connected source is approved before Pantasso reads it
- Cited answers with receipts from approved context
Company activity becomes governed operating context.
Pantasso does not stop at an answer index. It organizes approved source activity into governed company records, evidence, decisions, and work, and connects them through source-backed references. Canonical writes add validation, conflict handling, and transaction history while every record remains linked to its origin.
Structured workspace
- Typed company records with source-backed connections
- Full provenance on every record
- Queryable by SQL and natural language
- Transactional write boundaries for important state
- Incremental sync keeps objects current
Consistent current context for people, applications, and assistants.
Choose which sources may be used, require citations, define authority and freshness, and keep those rules consistent across your team, applications, and assistants.
Trusted answers
- Approved sources and information scope
- Required citations and receipts
- Freshness and response limits
- Repeatable quality checks
- Shared by people, apps, and assistants
Give AI assistants safe context and approved workflows.
Each AI assistant gets a profile that defines what it may access. Configured governed workflows add exact action permissions, approval requirements, and escalation rules without widening the assistant's source access.
Agent profiles
- Per-assistant source access controls
- Read, draft, and escalation boundaries
- Activity logs for every agent interaction
- Works with any LLM or agent framework
For developers and technical evaluators
Technical depth under a simple interface.
These are the platform layers Pantasso is building. In the public beta, start with files, cited answers with receipts, and founder-led first-source setup. Do not treat this list as a finished GA surface.
Transactional canonical objects
The platform is building validation, constraints, rollback, and audit-grade transaction logs for important company state. The public beta does not expose that write surface as a finished operator product.
Multi-model company data
The platform is building one governed model for records, documents, messages, files, and related objects. In the public beta, start with Files and Obsidian. Slack, Drive, Telegram, and GitHub need a connection before message or Drive objects appear.
Incremental sync
Connected sources are meant to sync incrementally after they are proven live. Files and Obsidian are the live first source. Slack, Drive, Telegram, and GitHub need a connection before incremental sync applies.
Workspace isolation
Each hosted workspace is a separate data boundary. Approved sources, objects, agent profiles, and query history are not shared across workspaces.
Audit and provenance
Cited answers in the public beta link back to approved Files and Obsidian records. Broader transaction, transform, and payload lineage is platform work still being built.
Backup and export
Hosted workspaces can be backed up and exported as structured data during founder-led evaluation. Support diagnostics are redacted before transmission.
Interoperability runtime
Typed connector contracts are being built for later families. A directory listing is not a live or GA connector. Slack, Drive, Telegram, and GitHub still need a connection.
Operational analytics
The platform is building operational projections, materialized views, and dashboards next to canonical state. Do not treat those surfaces as a public-beta product.
Deployment choice
The long-term platform includes hosted, dedicated, and operator-controlled environments. The public beta is a hosted files-first workspace, not a dedicated or customer-VPC product.
See what Pantasso can do with your company data.
See a cited answer in the live demo, then talk with us about the first source and question that matter to your company.