We're launching a new kind of data platform — one that understands processes, not just tables.
Today we're launching Sancalana. A data platform built around a simple idea: your business processes are the most important data you have, and SQL alone can't help you understand them.
SQL is powerful for answering questions about state: how many orders shipped last week, what's the average handle time, which region has the highest churn.
But it can't answer questions about flow: what path do orders actually take through your fulfillment process? Where do they get stuck? Why does the same process take 2 days for some customers and 2 weeks for others?
What SQL sees What you actually need
================ =========================
| order_id | status | Order 1: Created -> Approved -> Shipped
|----------|--------| Order 2: Created -> Review -> Approved -> Shipped
| 001 | Shipped| Order 3: Created -> Review -> Rejected -> Review
| 002 | Shipped| -> Approved -> Shipped
| 003 | Shipped| Order 4: Created -> Approved -> Delayed -> Shipped
| 004 | Shipped|
Same outcome. Very different journeys.
All four orders shipped. SQL says everything is fine. But orders 2, 3, and 4 took very different paths — and those paths tell you where your process is broken.
Sancalana takes your event logs and reconstructs the actual process:
Event Log Process Model
============ ====================
case activity time +----------+
---- ---------- -------- | Created |
001 Created 09:00 +----+-----+
001 Approved 09:15 |
001 Shipped 10:30 +----v-----+ +--------+
002 Created 09:05 | Approved +-->| Shipped|
002 Review 09:45 +----^-----+ +--------+
002 Approved 14:00 |
002 Shipped 15:30 +----+-----+
003 Created 09:10 | Review |
003 Review 11:00 +----------+
003 Rejected 13:00
... Reconstructed from 10,000+ events
From raw event data, our algorithms discover:
Discovery is just the start. Once you have the process model, you can:
Check conformance — overlay your intended process onto the discovered one. Where does reality deviate?
Analyze variants — why do 30% of cases take the review path? Is it correlated with order size, region, customer type?
Predict outcomes — based on where a case is right now, what's the probability it will be delayed?
We're shipping fast. On the roadmap:
| Feature | Timeline |
|---|---|
| AI-powered root cause analysis | March 2026 |
| Custom dashboards | March 2026 |
| Warehouse connectors (Snowflake, BigQuery) | April 2026 |
| Collaborative annotations | April 2026 |
Head to the platform page to learn more, or request early access — we'll map one of your processes live.