Holder’s seal

I help keep an enterprise lakehouse secure, governed and cost-transparent.

उपनाम / Surname: DUTTA
दिया गया नाम / Given name: BISHAKH
भूमिका / Role: Associate Software Engineer
जारीकर्ता / Issued by: Accenture, Azure Databricks
स्थान / Place: Asansol, India
से मान्य / Valid from: Jan 2026
मंज़ूरी / Clearances · core stack: Azure Databricks · Unity Catalog · PySpark · SQL · Azure Data Factory · Delta Lake · Entra ID & RBAC
प्राधिकरण / Authorisation: GRANT SELECT ON bishakh TO you;

Based in Asansol, India
IST (UTC+5:30)
Accenture since Jan 2026
B.Tech CSE 2025, CGPA 8.69

Clearances Selected work

Associate Software Engineer on Accenture's Azure Databricks platform team. I run Unity Catalog governance and least-privilege access, track what GenAI really costs, and build the pipelines that feed the lakehouse.

Unity Catalog access, governed as code

Type: Personal project
Year: 2026
Stack: Databricks SDK · Python · SQL · GitHub Actions

Access changes shouldn’t be clicks nobody can trace. I built a policy-as-code tool that applies Unity Catalog grants from version-controlled YAML through the Databricks SDK, so every change is reviewed in a pull request before it lands.

Drift detection flags excess or orphaned grants against the declared policy.
Row filters and column masks protect PII.
A Databricks SQL audit dashboard shows who accessed which data, and when.

Repository link: pending publication.

policy/finance.yaml

catalog: finance
grants:
  - principal: finance_analysts
    privileges: [USE_CATALOG, SELECT]
  - principal: finance_engineers
    privileges: [USE_CATALOG, SELECT, MODIFY]

drift report

✓ finance_analysts    matches policy
✓ finance_engineers   matches policy
✕ contractors_tmp     SELECT not in policy
  → flagged for revoke in PR #42

Illustrative example, not production data.

What GenAI really costs, per user and per workspace

Type: Accenture
Year: 2026
Stack: AI/BI Genie · Databricks SQL · Logic Apps

Teams were adopting Databricks Genie faster than anyone could see the bill. I built a usage tracker that follows token consumption and SQL warehouse usage per user and per workspace, so stakeholders can see adoption and cost side by side.

A weekly report goes out through Azure Logic Apps every Monday, with no manual effort.
An interactive Databricks dashboard drills down to each user and workspace.

Data flow

Genie token & SQL warehouse usage
Databricks SQL, per user & workspace
AI/BI dashboard with drill-down
Logic Apps report, every Monday

WorkspaceActive usersTokensWarehouse hrs

prod-analytics

finance-bi

data-science

Figures withheld: not yet cleared for publication. Workspace names are illustrative.

Everyone sees exactly what they’re cleared for

Type: Accenture
Scope: Workspaces across the platform
Stack: Unity Catalog · Entra ID · Azure RBAC · Key Vault

I map Entra ID groups and Azure RBAC to Databricks permissions, then let Unity Catalog row filters and column masks do the rest. Try it: switch the group below and watch the same query return different data.

Query as

-- attached once; the same query returns different data per groupCREATE FUNCTION crm.region_filter(region STRING)RETURN NOT is_account_group_member('east_analysts')OR region = 'East';ALTER TABLE crm.customers SET ROW FILTER crm.region_filter ON (region); CREATE FUNCTION crm.mask_email(email STRING)RETURN CASEWHEN is_account_group_member('platform_admins') THEN emailWHEN is_account_group_member('data_stewards')THEN regexp_replace(email, '^(..)[^@]*', '$1•••')ELSE '•••' END;ALTER TABLE crm.customers ALTER COLUMN email SET MASK crm.mask_email;-- phone uses the same pattern: only platform_admins see it unmasked

SELECT * FROM crm.customers; run as east_analysts. Highlighted policy lines decide this group’s result.

Name	City	Region	Email	Phone
Aarav Sen	Kolkata	East	masked	masked
Rohan Das	Bhubaneswar	East	masked	masked
Ishaan Ghosh	Asansol	East	masked	masked

3 rows hidden by the row filter.Synthetic data. Ink bars stand for masked values.

Travel history Experience & education

Where I’ve worked and studied.

Accenture
Associate Software EngineerJan 2026 – Present
Azure Databricks platform administration & data governance
- Run Unity Catalog governance: catalogs, schemas, table grants, and row- and column-level security so people see only the data they're authorised for.
- Enforce least privilege by mapping Entra ID groups and Azure RBAC to Databricks permissions; handle access provisioning and onboarding across workspaces.
- Moved secrets into Azure Key Vault-backed secret scopes, removing hard-coded credentials from notebooks and pipelines.
- Build and maintain ADF and Synapse pipelines into ADLS Gen2, PySpark jobs on Delta Lake, and SQL analysis that answers stakeholder questions.
CuriousJr
Software Developer Intern · RemoteOct 2021 – Jun 2022
Web apps and educational games
- Built web applications and educational games in JavaScript (Phaser, P2 physics) used in student coding courses.
- Earned an early promotion and mentored new interns.
Haldia Institute of Technology
B.Tech, Computer Science & Engineering2021 – 2025
CGPA 8.69 / 10

Endorsements Skills

The tools I’m cleared to work with.

Databricks

Unity Catalog: metastore, grants, row filters, column masks, lineage
Workspace administration
AI/BI Genie & AI/BI Dashboards
Databricks SQL
Delta Lake
Databricks SDK

Azure

Data Factory
Synapse Analytics
ADLS Gen2
Logic Apps
Entra ID
Azure RBAC
Key Vault

Languages

SQL
Python
PySpark
C++
JavaScript

DevOps & tools

Git & GitHub
GitHub Actions (CI/CD)
Jira

Request access

Hiring for data platform, governance or Databricks work? I’m based in Asansol, India and reply to every message.

bishakhdutta2002@gmail.com