I build data platforms that teams can trust.
Databricks data engineer with 5 years of experience across healthcare, finance and telecom.
I design batch and streaming pipelines on Databricks, Spark and Delta Lake. I govern them with Unity Catalog, ship them with Terraform and CI/CD, and evaluate the AI systems that run on top of them. I'm available for freelance data engineering and AI-training work in the U.S.
Reliable pipelines, governed data, measurable outcomes
I've built data platforms at Frontier Communications, Johnson & Johnson, Grapplesoft and Cloud Technologies. That work spans real-time ingestion with Kafka, Kinesis and Spark Structured Streaming, Delta Live Tables quality monitoring, and fine-grained access control for sensitive healthcare and financial data.
I care about the unglamorous parts that make data trustworthy: schema enforcement, lineage, quarantine tables, idempotent writes, cost-aware cluster sizing, and documentation teammates actually use. The three projects below show those habits end to end, on real U.S. government data, with dashboards you can use right now.
Lakehouse engineering
Databricks and Delta Lake medallion pipelines, DLT expectations, Unity Catalog governance, and Terraform with Asset Bundles.
Streaming & real-time analytics
Kafka, Kinesis and Spark Structured Streaming with exactly-once Delta writes, plus Java ingestion services.
AI training & LLM evaluation
Gold-answer datasets, rubrics and error taxonomies, SFT and DPO data, annotation QA, and evaluation harnesses.
Performance & cost tuning
Spark job optimization, Z-ordering and partition pruning, workload benchmarking, and cloud storage tiering.
Real U.S. data, real problems, live dashboards
Each project is a complete pipeline built on public U.S. government data. Click a card to open its interactive dashboard on its own page.
Rx Integrity: Medicare Part D lakehouse
Flags prescribers whose opioid and cost patterns sit far outside their specialty peers, and finds OIG-excluded prescribers who still had Medicare claims.
StormPulse: real-time weather exposure
Streams live National Weather Service alerts through Kafka and Spark to show ops teams how many people are under active weather warnings, with no double counting.
FinFacts-Eval: LLM financial-QA benchmark
Measures how LLMs misstate figures from 10-K filings: wrong units, wrong year, invented numbers. It turns every failure into labeled SFT and DPO training data.
Five years building production data systems
-
Jul 2025 – PresentUSA
AWS–Databricks Data Engineer
Frontier Communications- Integrated streaming ingestion with Kafka, AWS Kinesis and Databricks Structured Streaming for near-real-time reporting.
- Implemented Delta Live Tables quality monitoring (lineage, schema enforcement, alerts), cutting SLA breaches by 20%.
- Benchmarked workloads and recommended cluster sizing and caching that reduced compute costs by 15%.
- Worked with InfoSec on fine-grained access controls and encryption for sensitive healthcare data.
-
Dec 2024 – May 2025USA
Databricks Data Engineer Intern
Johnson & Johnson- Optimized PySpark and Scala ETL for high-volume financial transactions, 18% faster.
- Built Java microservices for real-time financial data validation and enrichment.
- Built Kinesis, Kafka and Structured Streaming (Scala) workflows for sub-second market-data analytics.
- Applied Unity Catalog policies and encryption standards for SOX compliance.
-
Apr 2022 – Aug 2023India
Data Engineer
Grapplesoft- Built DLT pipelines with lineage, schema evolution and audit logging, cutting data recovery time by 25%.
- Added schema-drift detection, null handling and duplicate resolution, giving 30% fewer downstream errors.
- Tuned Spark with Z-ordering, partition pruning and predicate pushdown, cutting unnecessary scans by 40%.
- Led internal "Data Engineering Guild" sessions on Spark debugging and cost optimization.
-
Jan 2021 – Mar 2022India
Azure Cloud Engineer
Cloud Technologies- Optimized Azure Data Factory pipelines from SQL Server, APIs and files, improving ETL performance by 20%.
- Cut combined Azure and S3 storage spend by 15%+ with lifecycle policies and tiering.
- Built Delta Lake pipelines on Azure Databricks with ACID transactions and time travel.
-
Jun 2020 – Nov 2020India
Cloud Engineer Intern
Intense Technologies- Provisioned Azure, AWS and GCP resources with ARM templates, CloudFormation and Terraform. Configured least-privilege IAM and RBAC.
Tools I use in production
Languages
Big data & lakehouse
Streaming
Cloud
Orchestration, IaC & CI/CD
Data stores, governance & BI
Let's talk about your data
Need a pipeline built, a Databricks workspace tuned, or an LLM evaluated against real-world data? Tell me about the project. I usually reply within one business day.