Problems I Solve
Reduce policy drift in distributed data stacks by unifying metadata quality, policy generation, and query-time enforcement.
Cybersecurity Platform Engineering
Vivek Jain
I design distributed data and policy systems that solve real cybersecurity challenges at scale: identity-aware access, governed data sharing, and low-latency authorization.
- Distributed system design for cyber operations
- Metadata-to-policy automation
- Data quality + agentic governance pipelines
Cyber + Distributed Systems Profile
I build resilient platform control planes where distributed data execution and cybersecurity governance are engineered as one system.
Security Outcome
Consistent authorization behavior across services, stronger audit trails, and safer scaling for platform and security teams.
Featured Project: federated-quasar
A distributed cybersecurity data operating model that turns metadata into enforceable policy across multiple control planes.
What It Delivers
- Executes distributed dbt workflows against Trino via Governance Execution Engine
- Publishes technical and business metadata into OpenMetadata as a shared source of truth
- Adds data quality checks and scorecards into the delivery path
- Transforms metadata into security policy bundles through Moat + OPA
- Enforces low-latency authorization decisions directly in Trino
4
Distributed namespaces
7+
Security-aware systems integrated
1
Governed cyber execution path
100%
Traceable flow with run logs
Agentic MCP + Data Quality Layer
On top of Trino, I build an MCP-powered agentic layer that joins execution context from Trino MCP and metadata context from OpenMetadata MCP.
Joined MCP Architecture
- Trino MCP exposes query, schema, and execution context for AI agents
- OpenMetadata MCP exposes lineage, ownership, tags, and glossary context
- Federation layer joins both MCP outputs into one policy and quality context
- Agentic planner uses joined context to generate safe SQL and governed actions
2
MCP sources joined
1
Federated decision context
DQ
Data quality gates before publish
RAG
Agentic metadata retrieval path
Trino MCP
Capture runtime context, query patterns, and table-level execution metadata.
OpenMetadata MCP
Capture lineage, ownership, tags, and semantic metadata for governance reasoning.
Context Join Layer
Join Trino MCP + OpenMetadata MCP payloads into a single policy-aware context.
Data Quality Gate
Apply freshness, completeness, and validity checks before agent-led actions.
Agentic Planner
Use joined context to generate safe SQL/API actions with policy constraints.
Governed Execution
Execute through Trino with auditable decisions, policy checks, and quality evidence.
End-to-End Cyber Governance Flow
From distributed model execution to runtime policy decisions in one connected sequence.
Load Workload
Start a governed run in GEX with model selection and execution control.
Execute Distributed Query Path
Run dbt run/test/docs in Trino to create and validate shared data assets.
Ingest Metadata + Data Quality Context
Push Trino and dbt artifacts into OpenMetadata with data quality checkpoints.
Generate Cyber Policies
Moat maps metadata attributes into policy bundle state.
Distribute to OPA
OPA consumes bundles and serves low-latency decisions in real time.
Enforce at Runtime
Trino evaluates allow, row-filter, and column-mask decisions for every query.
More Cyber Platform Work
Additional projects and writing on policy engineering and secure data operations.
Kubernetes Components POC
Separate page for component areas, deployment snippets, and direct codebase links.
Open components pageRego Policy Studio
Policy authoring and validation workflows for governance teams.
Open repositoryMedium Articles
Full story list with links, read time, publish dates, and performance metrics.
Open article pageLegacy Tools Archive
Earlier calculators and utilities are preserved in the archive section.
Open archive