GitHub ↗ LinkedIn ↗ Contact me
OPEN TO SENIOR DATA ENGINEERING ROLES

SENIOR DATA ENGINEER · RALEIGH, NC

I build cloud data platforms for batch and streaming workloads.

I’m Bhanu Vejendla, a Senior Data Engineer with 7+ years of experience turning high-volume data into reliable systems across Azure, AWS and GCP.

Explore my projects ↓ Discuss a role ↗

Review my professional experience ↓

platform_overview.yaml ILLUSTRATED PIPELINE
INGESTION BATCH + STREAM QUALITY CONTRACTS + TESTS OPERATIONS MONITORED + RECOVERABLE
EXAMPLE FLOW
INGESTING EVENTS
SOURCES READY
API / CDC
EVENTS / FILES
INGEST READY
KAFKA
AIRFLOW / ADF
PROCESS READY
SPARK
DATABRICKS
STORE READY
DELTA LAKE
SNOWFLAKE
SERVE READY
BI / ML
DATA PRODUCTS
schema contracts quality gates lineage & observability failure recovery
7+ years engineering
data platforms
3 cloud ecosystems
Azure · AWS · GCP
5 featured engineering
projects
4 industry domains
Insurance · Gaming · Financial Services · Healthcare
01 · ABOUT

I care about what happens after a pipeline goes live.

A good data platform should be reliable today and understandable six months from now.

Bhanu Vejendla

I started my data engineering career in India and later continued it in North Carolina, moving through healthcare, financial services, gaming and insurance. Those environments were very different, but they taught me the same lesson: data becomes valuable only when people can trust it.

That is why I care about the work after launch—the quality checks, recovery paths, monitoring, security, performance and cost that keep a platform dependable. I build reusable patterns so the next engineer can understand, operate and extend the system without starting over.

02 · EXPERIENCE

What I’ve worked on.

Each role has strengthened a different part of my data engineering experience.

METLIFE · CARY, NORTH CAROLINA

Azure Data Engineer

Reusable ingestion patterns and reliable reporting datasets.

  • Build Azure Data Factory and Databricks pipelines from ADLS Gen2 into Synapse and Azure SQL.
  • Process terabytes of data with Python and Spark SQL to support customer reporting.
  • Develop reusable Python patterns for change data capture, deduplication and incremental loading.
  • Monitor dataset freshness, triage failed loads and apply masking and role-based access for sensitive data.
EPIC GAMES · CARY, NORTH CAROLINA

AWS Data Engineer

90% less infrastructure management time through Terraform automation.

  • Developed AWS Glue and Spark workflows to move campaign data from S3 into Redshift.
  • Added metadata validation, reconciliation and error handling to protect reporting accuracy.
  • Tuned Snowflake queries and built Elasticsearch/Kibana monitoring for transaction visibility.
INCRED FINANCIAL SERVICES · MAHARASHTRA, INDIA

GCP Data Engineer

Curated BigQuery datasets and visibility into platform costs.

  • Built Airflow pipelines on GCP and PySpark notebooks for business data models.
  • Processed and monitored workloads across BigQuery, Dataproc and Dataflow.
  • Created billing and usage reports to support query optimisation and cost decisions.
SIEMENS HEALTHINEERS · MAHARASHTRA, INDIA

Data Engineer

Batch and streaming pipelines with built-in quality checks.

  • Developed warehouse ETL with Python, Snowflake and SnowSQL.
  • Integrated Kafka with Spark Streaming and provisioned Databricks workloads.
  • Implemented validation, anomaly detection and profiling, with Airflow and ADF scheduling.
03 · TECHNICAL STACK

Tools I’ve used to deliver data systems.

Grouped by where they fit in the work, from ingestion through production delivery.

01
INGESTION & ORCHESTRATION

Reliable movement from source to platform.

Kafka Airflow Azure Data Factory APIs & Postman Apache NiFi AWS Glue Cloud Dataflow Apache Beam Java Message Service
02
PROCESSING & MODELING

Batch and streaming transformations at scale.

Apache Spark Databricks Python & PySpark dbt SQL Hadoop Apache Storm Dataproc Java
03
LAKEHOUSE & WAREHOUSE

Governed storage and analytics-ready models.

Snowflake ADLS Amazon S3 BigQuery Delta Lake Apache Parquet PostgreSQL Azure Blob Storage Azure Synapse Azure SQL Warehouse Amazon Redshift DynamoDB Elasticsearch SnowSQL
04
CLOUD, PLATFORM & DELIVERY

Repeatable infrastructure and production delivery.

Microsoft Azure AWS Google Cloud Terraform Docker Kubernetes GitHub Actions Jenkins Git Grafana Kibana AWS Lambda Cloud Monitoring Looker Studio Docker Swarm
ENGINEERING PRINCIPLES

Idempotent pipelines · Data contracts · Automated testing · Checkpointing · Backfills · Lineage · Monitoring · Secure access

04 · ENGINEERING WORKFLOW

How I take a pipeline from idea to reliable operation.

The tools change by platform. The engineering discipline stays consistent.

delivery_workflow.yaml DISCOVERING SOURCES

Illustrated workflow showing my delivery process.

EXAMPLE STATE HEALTHY EXAMPLE QUALITY PASSING RECOVERY IDEMPOTENT
  1. 01 · DISCOVER

    Understand the source

    Volume, latency, ownership and failure modes.

  2. 02 · CONTRACT

    Define expectations

    Schemas, quality rules, SLAs and access.

  3. 03 · INGEST

    Move data safely

    Replayable batch or streaming ingestion.

  4. 04 · TRANSFORM

    Build trusted models

    Tested, incremental and scalable processing.

  5. 05 · DELIVER

    Serve the right shape

    Governed tables and analytics-ready products.

  6. 06 · OPERATE

    Design for reality

    Monitoring, recovery, performance and cost.

✓ Schema contracts ✓ Quality checks ✓ Lineage ✓ Observability ✓ Failure recovery
05 · SELECTED ENGINEERING WORK

Projects that show how I work.

Five projects showing production ownership, practical architecture and reliability—from a live community platform to data engineering systems.

Choose a project to review the problem, engineering decisions and supporting evidence. Each case study links to the live site or source repository.

FLAGSHIP PRODUCTION PLATFORM

Kingdom846

Kingdom846 public homepage showing the royal portal, alliance navigation and community features
SITE USAGE
✓1,719 visits✓5,968 pageviews✓58 countries✓3.5 pages per visit
PROBLEM

A gaming community needed one dependable place for alliance information, live rosters, applications, rankings, voting and operational updates. I built Kingdom846 as a responsive platform used by real community members.

ENGINEERING DECISION

I separated the React interface from an Express API, persisted operational data in SQLite and enforced role boundaries for administrators, alliance leaders and War Room users. Rate-limited authentication, version-aware updates and a server-side Royal Advisor make the platform practical to operate.

DECISION UNDER PRESSURE

Preventing one teammate from silently overwriting another

01
The failure mode

Several authorised War Room users can edit one shared roster. A simple last-write-wins save could erase newer work without warning.

02
The safeguard

Each client saves the version it started from. The API refuses a stale version with a conflict response instead of accepting an unsafe write.

03
The recovery path

The interface shows who saved the current version and lets the editor load it or deliberately replace it. The conflict becomes visible and recoverable.

ARCHITECTURE FLOWILLUSTRATED WORKFLOW
01 · EXPERIENCEReact + ViteResponsive portal
02 · SERVICESExpress APIRole-based access
03 · DATASQLiteLive community state
04 · OPERATEProduction siteReal users + updates
Role-based accessConflict-safe writesRecoverable editing
ReactExpressSQLiteREST APIs
Visit live website ↗
ENGINEERING NOTES · PROJECT DECISIONS

The reasoning behind the systems.

Five short technical stories about failures, trade-offs and safeguards that are difficult to show in a project summary.

01 · KINGDOM846

When two teammates save at the same time

Turning a silent last-write-wins risk into a visible, recoverable conflict.

Express · SQLite · Concurrency 02 · WEBLOG STREAM

Exactly-once is won in the sink

How a count exposed silent data loss and led to replay-safe Parquet writes.

Kafka · PySpark · Parquet 03 · NYC TAXI LAKEHOUSE

A batch is finished only after its output commits

Making monthly loads safe to retry with partition replacement and a durable manifest.

PySpark · Manifests · Backfills 04 · RETAIL WAREHOUSE

Late data is why incremental models need a lookback

Reprocessing a small trailing window without duplicating yesterday’s orders.

dbt · Incremental SQL · Testing 05 · DATA PLATFORM TERRAFORM

Asking for force-destroy is not the same as receiving it

Encoding production safety in the configuration and proving it at plan time.

Terraform · AWS · Guardrails
LET’S TALK · SENIOR DATA ENGINEERING

Let’s talk about your next data engineering hire.

Building a team around batch pipelines, streaming systems or cloud data platforms? Share the role, location and the engineering challenges you’re working on.

Email me about a role Connect on LinkedIn

bhanumanideepvejendla@gmail.com