Shield Data Map

Discover every personal data column. Automatically.

Deploy a lightweight agent on your infrastructure. It scans your databases and filesystems, classifies PII against DPDPA categories, and feeds your RoPA — without raw data ever leaving your servers.

Why you need it

DPDPA S.8(3) requires you to erase personal data once its purpose is fulfilled. You cannot erase what you have not mapped.
Manual RoPA exercises miss PII in logs, exports, and legacy tables. Discovery finds what you forgot you had.

Raw data stays on your servers. Only findings metadata (column name, PII type, confidence) reaches the platform. Verified in the agent source code.

How It Works

From zero to a complete personal data map in under an hour

1
Deploy an agent
Download the Docker image from the dashboard and run a single command to register the agent on your own infrastructure. It stays inside your network.
2
Connect your sources
Configure read-only credentials in scan.yaml for your databases (PostgreSQL, MySQL, MongoDB, MSSQL) and filesystem directories.
3
Agent scans locally
The agent connects with read-only access, discovers every table and column, and classifies PII using pattern matching. Raw values are discarded after classification — only metadata is sent to the platform.
4
Map is built automatically
Findings appear as classified assets tagged with DPDPA PII tiers (Special Category, Sensitive, Personal). Link them to processing purposes with one click.
5
RoPA and breach scope are ready
Promote discovered assets into your RoPA with pre-filled data. When a breach occurs, instantly scope which columns — and how many individuals — were potentially exposed.
Features

Built for security-conscious organisations

Core

Agent runs on YOUR infrastructure

The scanner agent is a Docker container you deploy inside your own network. It makes outbound HTTPS calls to the platform only — no inbound ports required.

Raw data never leaves your servers

Column samples are analysed locally and discarded immediately after classification. What reaches the platform is metadata only: column name, PII type, confidence score, row count.

5 connector types

PostgreSQL, MySQL, MongoDB, Microsoft SQL Server, and filesystem (CSV, JSON, Excel, logs). Read-only enforcement at the driver level — the agent cannot write to your databases.

DPDPA-aligned PII classification

40+ Indian PII types mapped to DPDPA tiers: Special Category (Aadhaar, PAN, health data), Sensitive (financial, biometric), and Personal. Powered by pattern matching and optional AI reclassification.

One-click RoPA promotion

Discovered assets are auto-matched to your existing processing purposes. Review candidates in the Operations tab and promote them to your RoPA with pre-filled metadata.

Breach blast-radius scoping

Link scanner sources to breach incidents. Instantly see which PII categories and how many data principals were potentially exposed — with the exact table/column breakdown.

Anti-tamper hardening

Each agent instance is bound to its host environment via a cryptographic fingerprint. Copying an agent credential to another machine is detected and flagged as a security alert.

Scheduled + on-demand scanning

Configure daily, weekly, or monthly scans per source. Trigger manual scans from the dashboard. Drift alerts notify you when new PII columns appear or existing ones disappear.

5
Connector types
40+
Indian PII types
~20 MB
Idle RAM footprint
100%
Data stays on-prem
Available from Starter plan onwards

Shield Data Map is available on all plans. Scan frequency and number of sources vary by plan.

FAQ

Common questions

Why does DPDPA require automated data discovery?+

DPDPA Section 8(3) requires Data Fiduciaries to erase personal data once its processing purpose is fulfilled. You cannot erase what you have not mapped. Many organisations find PII in unexpected places — logs, exported CSVs, legacy tables — that manual RoPA exercises miss entirely. Shield Data Map answers the first question: what personal data do you actually hold, and where?

Does the agent read my actual data?+

The agent samples up to 500 row values per column for pattern matching. These values are analysed locally inside your infrastructure and discarded immediately after classification. The platform only receives metadata: schema name, table name, column name, detected PII type, confidence score, and row count estimate. No cell values, no hashed identifiers, no raw data ever leaves your servers.

What databases and sources does Shield Data Map support?+

The slim image supports PostgreSQL, MySQL, MongoDB, and filesystem sources (CSV, JSON, Excel, log files up to 50 MB). The full image adds Microsoft SQL Server via ODBC Driver 18. Oracle, SAP HANA, Snowflake, BigQuery, and cloud-native stores are on the roadmap but not yet available.

What does "read-only" mean in practice?+

For PostgreSQL, the agent sets SET SESSION CHARACTERISTICS AS TRANSACTION READ ONLY at the driver level and verifies it before scanning — a write attempt would throw an error. For MySQL, it checks SHOW GRANTS and aborts if INSERT/UPDATE/DELETE permissions are present. Filesystem mounts use the :ro Docker flag, enforced at the kernel level.

Which plan includes Shield Data Map?+

Shield Data Map is available on all plans including Starter. The number of sources and scan frequency vary by plan. See pricing for details.

Know exactly what personal data you hold

Deploy the agent, connect your databases, and get a complete personal data map in under an hour. Your data never leaves your infrastructure.