Simorgh
Connect your database. In V2, Simorgh stops waiting to be asked —
and starts telling you what matters before you think to look.
Closed beta — currently onboarding select business partners
Your data team answers questions.
Nobody is asking the right ones.
Every week, critical signals are buried in your database. Revenue is shifting. Churn is building. An opportunity is closing. But your executives are waiting on a report, your analysts are writing SQL, and by the time the answer arrives — the moment has passed.
The problem isn't access to data. It's the time between something happening and someone understanding it.
Ask your database anything.
Get a real answer in seconds.
No new tool. Plug into the LLM you already use.
Simorgh reads your database schema and builds MCP servers that expose your data as tools — directly inside Claude, or any LLM your team already works with. No switching platforms. No new interface to learn. Your analysts keep using Claude. Your database just becomes visible to it.
Native Claude integration included. Works with any MCP-compatible AI — the same connection layer, any model.
Claude · Any MCP-compatible LLM · Read-only · No migration · Privacy mode available
Governed tools, reviewed by your team.
We compile your data model into a set of governed, approved query tools — specific to your business, your entities, your metrics. Every tool generated from your schema goes through a human review gate before it can execute. Nothing runs without your sign-off.
Approved by your team · Business context layer · Your formulas, enums, vocabulary
Plain English. Real answers. In seconds.
Your executives, analysts, and operators ask questions in plain English. Simorgh selects the right tools, executes the query safely, and returns a plain-English answer with the data behind it. Any question that takes hours today — answered in seconds.
Full audit log · Multi-tenant isolation · Source data always visible
Try it. Ask anything.
Watch Simorgh answer real business questions in real time.
Read-only connection · No data migration · Privacy mode available
The governed intelligence layer your data team can trust.
Natural Language Q&A
Ask any question in plain English. Simorgh selects the right query tools, executes safely, and returns a human-readable answer with source data.
Auto-Generated MCP Tools
We compile your schema into governed query tools specific to your business entities and metrics. Every tool reviewed and approved by your team before use.
Business Context Layer
Teach Simorgh your formulas, enums, and vocabulary. It learns the difference between "active customer" and "churned customer" the way your team defines them.
Read-Only Enforcement
Your database user is verified to have zero write permissions — at the credential level, not just query-level blocking. Your data is structurally protected.
Privacy Mode
Run Simorgh inside your own network. Your credentials never leave your infrastructure. We only receive query results — never raw data.
Full Audit Log
Every query, every tool call, every AI response is logged with org, user, timestamp, and token usage. Full accountability for business compliance.
Any question your leadership asks today that takes hours or days to answer — answered in seconds, with the source data visible.
This is not another dashboard.
This is not a data chatbot.
| Traditional BI | Data Chatbot | Simorgh | |
|---|---|---|---|
| Requires you to ask | —Always | —Always | ✓No — proactive in V2V2 |
| Understands your business | —No | —Partially | ✓Business Model GraphV2 |
| Explains root cause | —No | —No | ✓Investigation EngineV2 |
| Gets smarter with use | —No | —No | ✓Compounding memoryV2 |
| Works on your existing DB | —Partially | —Partially | ✓Yes — no migration |
| Privacy / on-premise mode | —Varies | —Varies | ✓Yes |
| Full audit log | —Partial | —Rarely | ✓Every query logged |
| Human approval gate | —No | —No | ✓Every tool reviewed |
V2 = Roadmap — currently in development
Governed. Audited.
Read-only. Yours to control.
Read-Only at the Credential Level
We verify your database user has zero write permissions — not just query-level blocking. Your data cannot be modified. Ever.
Nothing Runs Without Approval
Every tool generated from your schema passes through a human review gate before it can execute. Your team signs off on every query path.
Privacy Mode Available
Run Simorgh's server inside your own network. Your credentials never leave your infrastructure. We only receive query results — never raw data.
Full Audit Log
Every query, every tool call, every AI response is logged with org, user, timestamp, and token count. Complete accountability for compliance teams.
Multi-Tenant Isolation
Your data is scoped and inaccessible to other organizations at the database level. Business-grade isolation from day one.
No Vendor Lock-In
Connects to your existing Postgres database. No data migration, no new warehouse, no proprietary format. Leave whenever you want.
V1 answers your questions.
V2 asks better ones than you
would have thought to.
The current version of Simorgh is the foundation. A governed, intelligent query layer your team can trust. But the real shift is what comes next.
In V2, Simorgh stops being reactive. It builds a living model of your business — your revenue streams, your customer lifecycle, your KPIs — and runs continuously in the background, watching for changes, testing hypotheses, and ranking what it finds by business impact. When something matters, Simorgh tells you unprompted.
Continuous Insight Engine
Simorgh monitors your KPIs on a configurable cadence. Trend shifts, anomalies, and cohort divergences are detected automatically — no dashboard to check, no alert to configure.
Change Decomposition
When a metric moves, Simorgh doesn't just tell you it moved. It tells you why — decomposed by customer tier, region, product, cohort, and time period. Ranked by contribution magnitude.
Hypothesis Engine
On initialization, Simorgh generates a library of hypotheses specific to your business type. It tests them continuously. Confirmed correlations build into a permanent Business Model Graph.
Investigation Engine
Every significant finding follows a structured chain: Observation → Validation → Root Cause → Business Impact → Recommendation. The evidence is always shown.
Executive Intelligence Layer
Raw findings become executive narratives. Not "revenue dropped 12%" — but a full explanation of what drove it, which segment, which cause, and what action to take.
Insight Feed
A continuously updated, ranked, stakeholder-filtered feed of what matters. Risk alerts. Opportunity signals. Weekly reports. Delivered to inbox, Slack, or dashboard.
The system gets smarter over time.
Every piece of feedback — correct, incorrect, irrelevant, important — updates the ranking model and strengthens the business memory graph. The longer Simorgh runs, the more precisely it understands your business.
Performance your analysts will notice.
Results your executives will demand.
From hours to seconds.
Real answers, not batch jobs.
On your governed query tools.
See Simorgh working
on your data.
We are onboarding a small number of businesses in our closed beta. If you are building toward a future where your leadership has real-time intelligence instead of delayed reports — we would like to show you what that looks like.
We will respond within 24 hours. No commitment required.