Analytics and AI that run where your data lives.
A family of pre-engineered components that compose into one platform: workflows, an orchestration engine and an analytics module where third parties collaborate inside your boundary. All of it runs on Zero Trust Orchestration; the work runs beside your databases.
SOC 2 Type II attested 2023–2025 · ISO 9001 · Zero Trust Orchestration, patent-pending
Start with the component your first workload needs.
KEZEL is a family of pre-engineered software components built over the Zero Trust Orchestration concept. Adopt one and it works alone; together they run your workflows, orchestration and analytics.
Workflows
Agent workflows, ETL and scheduled jobs that act with permission. Every action is policy-checked, and every run lands on the ledger.
Orchestration engine
The Zero Trust Orchestration runtime underneath everything: policies compiled into encrypted instructions, dispatched to agents beside your data.
Analytics module
Dashboards (VizEngine), natural-language queries (LLMCore) and the Insights Studio engines, over governed views. Third parties collaborate on those views inside your boundary, never on the records.
pre-engineered by DBTEZ · adopt independently, compose into one platform · every component runs on Zero Trust Orchestration
How a request runs without a copy.
The orchestration engine at the family's core implements Zero Trust Orchestration, DBTEZ's patent-pending architecture. The four beats, in order:
Policy
Your rules, versioned: row filters, column masks, aggregation floors, reader roles.
Encrypted instruction
Built from schema metadata and intent. The only thing that crosses the boundary, inbound.
Compute
The agent does the work beside your tables: queries, transformations, model runs.
Read under policy
Results stay where they were made, read through the policy gate. Every read is logged.
boundary crossings · one, inbound, encrypted · consumption · in-boundary reads through the policy gate
Engines for the questions a business repeats.
KEZEL Insights Studio, introduced at Bengaluru Tech Summit 2025 and carried by the analytics module. Each engine answers beside the system of record.
“How much?”
regression
“Which ones?”
classification
“When?”
time series
demand · actuals then projection · sample data
Every sector page is written as a deployment.
The lifecycle it serves, the question asked, where the agent sits, which components carry it. Most of these workloads also rotate through the console above.
Recent writing from the team.
Positions on the architecture, with the trade-offs conceded in writing. Figures drawn on sample data.
Bring a real workload. We’ll answer it in place.
Start with a demo on our sample boundary. Move to a pilot in yours once your security review clears it.