Vedhraby Ken42
Get started

Your environment. Your rules. Your data stays yours.

Vedhra runs where you choose, connects to every system you run, and governs what is known and what is done.

Two controls

Control over what is known, and what is done.

Knowledge controls

policy.read
01 What can be known
02 By whom
03 Under which definition
04 From which source

Action controls

policy.act
01 What can be done
02 By whom or by which agent
03 Under which conditions
04 With which approval
01

Permission follows context.

Access is applied where knowledge is served, so every interface inherits the same rules.

02

Every fact has provenance.

Any answer traces back to its source record and transformation.

03

Every action has authority.

People, applications and agents act within explicit permissions.

04

Every change leaves a trail.

Decision, approval, execution and outcome are recorded and auditable.

Access

MCP, SQL and API, through one governed door.

The same definitions and the same row-level policy apply to all three.

Mount a governed workspace over MCP. Tools are typed and scoped to the permissions you grant.

vedhra clients add leadership-assistant \
  --workspace university \
  --scopes metrics.read,entity.read,docs.search
 
# => endpoint  https://<your-workspace>/mcp
#    tools     governed metrics, resolved entities,
#              document search
Identity resolution

Five records, one person, with the evidence kept.

Five records, five systems
Admissions email · exact Ananya Rao · ananya.rao@gmail.com
Student Information System enrolment no. · exact Rao, Ananya · 24BBA0187
LMS email · exact user arao24
Finance name + programme · 0.94 Rao A. - BBA 2024
HR reviewer · confirmed A. Rao · teaching assistant
vedhra_id
vd_8f41c2
Ananya Rao · BBA · 2024 intake
one person record · email exact · enrolment no. exact · name and programme 0.94 · confirmed by reviewer
Illustrative data
Connectors

New connectors follow the same governed pattern, from contract to Gold.

Salesforce CRM
Microsoft Dynamics CRM
Moodle LMS
Canvas LMS
DBD2L Brightspace LMS
TPTally Prime Accounting
Microsoft 365 Identity and mail
Google Workspace Identity and mail
Zoho CRM and applications
DDarwinbox HRMS
Razorpay Payments
SAP ERP
Oracle ERP
How data comes in

Every source syncs into your own data lake.

Each sync copies only what is new or changed since the last run, into a data lake in your environment.
Vedhra reads from that lake. It never runs queries against your live systems.

1Your systems
Salesforce
Moodle
SAP
TPTally Prime
SISStudent Information System
XLSpreadsheets
DBExternal databases
DODocuments
2Incremental sync

Only new and changed records move, on a schedule set per source.

✓APIs and change logs, every 15 minutes to hourly
✓Scheduled batch for files and data dumps, nightly
✓A small agent beside on-premise software, connecting outward only
3Your data lake

In your cloud account or a dedicated in-region environment, in open formats you own.

Rawexact copy, append-only
Bronzetyped
Silvercleaned, deduplicated
Canonicalone model, identity resolved
Goldbusiness-ready
4Vedhra

Intelligence, agents and every interface read from the lake, under one set of permissions.

✓No load on your live systems
✓Every answer traces to a lake record
✓Approved actions are written back through the owning system
Data protection and sovereignty

Built for how institutions must handle data.

Data residency

Stays in-region, inside your own environment.

DPDP Act

Purpose-bound access and erasure across every copy.

Model independence

Approved models only. Your data never trains them.

Credentials

Encrypted, sealed and rotated on a schedule.

Audit

Every read and action logged with caller and purpose.