Question 1

What is an Organisational System of Record (OSR)?

Accepted Answer

An Organisational System of Record is the verified intelligence layer that sits on top of every system you already run — SAP, Tally, Salesforce, Workday, Oracle, your data warehouse, your docs. Instead of replacing them, the OSR ingests their events, resolves who-did-what-when across them, and exposes a single grounded interface for memory, workflows, process mining and AI agents. Crestline AI is the first OSR built for the modern enterprise.

Question 2

How is Crestline AI different from Glean, Notion AI or Microsoft Copilot?

Accepted Answer

Glean and Copilot are search-and-summarise tools over documents. Crestline is an operational layer with a verified org graph and a 4-layer memory model (user, project, department, organisation). It does not just retrieve text — it knows which person reports to whom, which task is blocked by which decision, which workflow ran when, and it grounds every answer with citations to the underlying event log.

Question 3

How is Crestline different from Celonis for process mining?

Accepted Answer

Celonis is a heavyweight enterprise process-mining platform with long implementation cycles. Crestline ships process mining as Module 13 — a CSV / Postgres / Salesforce / SAP / Workday / Oracle / Tally stream becomes a discovered process graph in minutes, with bottlenecks pulsed in red and quantified in rupees per year. Because the org graph already knows who handled each event, the bottleneck attribution is more specific (down to the named actor and department).

Question 4

How is Crestline different from n8n, Zapier or Make for workflows?

Accepted Answer

n8n and Zapier wire APIs together with a visual builder. Crestline workflows are authored in natural language, validated against the live org graph, and use org-native tokens like @manager_of(@self) or @cxo that resolve dynamically as people move between teams. The workflow engine runs through a five-stage agent pipeline — intent, planner, drafter, validator, narrator — so every step is explainable and auditable.

Question 5

Does Crestline replace SAP, Tally, Salesforce or my data warehouse?

Accepted Answer

No. Crestline is explicitly NOT a replacement for systems of record like SAP, Tally, Salesforce, Workday or Snowflake. It sits on top as the AI-native operating layer — ingesting events, building the verified org graph, and serving the AI Hub, Workflows, Console and Process Mining over the data those systems own. Think of it as the SAP of the modern world: not another silo, but the layer that makes the silos legible.

Question 6

What is the 4-layer memory model?

Accepted Answer

Every query routes through four parallel memory layers — USER (your personal context), PROJECT (active initiatives you participate in), DEPARTMENT (the team you sit in), and ORGANISATION (the whole company). The AI Hub fans the query out across all four in parallel, ranks matches by relevance and recency, and synthesises a grounded answer. This is why Crestline answers stay correct as people move between projects and teams — the memory compounds at every layer.

Question 7

Is Crestline AI secure and DPDP compliant for Indian enterprises?

Accepted Answer

Yes. Crestline is built India-first with DPDP compliance as a baseline — every event is consent-logged, the Constitution module hash-chains every org-document revision for tamper-evident audit, and SOC 2 Type II controls cover the platform. Data residency, role-based access, and per-tenant encryption are standard. Enterprise customers can deploy in a dedicated tenant in Indian regions.

Question 8

Who is Crestline AI built for?

Accepted Answer

Crestline is built for organisations of 100 to 5,000 people whose CFO, COO or CHRO needs to see what is actually happening across departments — bottlenecks costing real money, decisions buried in meetings, processes that drift from policy. It is most valuable where data already lives in SAP, Tally, Salesforce, Oracle or Workday but no human can hold the whole picture in their head.

Question 9

How long does Crestline AI take to deploy?

Accepted Answer

A typical Crestline deployment goes from connector setup to first quantified bottleneck in under a week. The CSV connector works in minutes; Postgres, Salesforce, SAP, Workday, Oracle and Tally connectors are standard. The verified org graph and 4-layer memory populate as your existing event streams flow in — there is no rip-and-replace and no upfront data-engineering project.

Customer health, in full context.