Software Engineer III, Finance Technology | Distributed systems, data platforms, and event-driven financial workflows
Portfolio | LinkedIn | LeetCode | Email
I build production systems where correctness, scale, and operational clarity matter. My current work sits at the intersection of finance technology, distributed backend systems, event streaming, cloud data platforms, and reconciliation workflows at Walmart Global Tech.
- Languages: Scala, Python, SQL
- Big Data and Streaming: Apache Spark, Spark Streaming, Kafka, KCaaS, Airflow, ETL Pipelines, Data Modeling
- Cloud and Infrastructure: Google Cloud Platform, BigQuery, Pub/Sub, Cloud Storage, Storage Transfer Service, AutoClass, Scheduled Queries, Kubernetes, ArgoCD
- Backend and Systems: Microservices, Distributed Systems, Event-Driven Architecture, REST APIs, PostgreSQL, MySQL, Flyway
- Testing and Delivery: ScalaTest, Mockito, Integration Testing, Litmus, CI/CD
- AI and Knowledge Systems: LLMs, RAG, Vector DB, Graph DB, ChromaDB, NetworkX
- Scala Ecosystem: ZIO, Akka/Pekko, Apache Camel
| Project / Initiative | What changed |
|---|---|
| Cloud Storage Archival Framework | Designed and owned GCS archival for 13+ month-old production data, moving 104 TB to archival storage and reducing yearly cloud cost. |
| Audit Event Listener Enhancement | Built REST APIs and parallel event processing for large backlogs, cutting event trigger latency by 2-3 seconds and reducing 1,000-message backlog processing by about 3 hours. |
| Ending Inventory Reconciliation | Contributed to event-driven, line-item-level reconciliation and backward acknowledgements, moving finance workflows from batch toward near real time. |
| Recon Adapter Enhancements | Built unified reconciliation capabilities that improved failure visibility, reduced manual troubleshooting, and lowered MTTD/MTTR. |
| Lead Company Code Derivation | Implemented 7 derivation strategies to improve tax accuracy and compliance in finance workflows. |
| GLAD PySpark Upsert | Optimized DataFrame upserts into PostgreSQL, reducing runtime from roughly 3 hours to 13 minutes. |
| TwinAI | Designed a passive knowledge digital twin concept that captures IDE, terminal, browser, meeting, and LLM context into queryable knowledge bases with privacy guardrails. |
| CodeSage AI | Built an LLM/RAG wiki pipeline that converts raw documents into linked, human-readable organization knowledge pages. |
| Period | Role |
|---|---|
| May 2026 - Present | Software Engineer III, Walmart Global Tech - Finance Technology, CILL |
| Jul 2024 - Apr 2026 | Software Engineer II, Walmart Global Tech - Finance Technology, CILL |
| Jan 2024 - Jun 2024 | Software Engineering Intern, Walmart Global Tech - Finance Technology, CILL |
| May 2023 - Jul 2023 | Software Development Engineer Intern, Walmart Global Tech - SAP Accounts Receivable |
Ask me about:
- Designing reconciliation systems where missing one line item is a business problem, not just a technical bug.
- Turning a multi-hour PySpark/PostgreSQL job into a minutes-level workflow.
- Cost-aware cloud data movement in GCS and BigQuery.
- Building event-driven source-to-ledger-to-SAP pipelines.
- How I would design an internal knowledge twin for engineering teams without ignoring privacy boundaries.
- Portfolio for the polished walkthrough: anirudh-vadera.github.io/Portfolio2023-24
- GitHub repositories for public project code: github.com/ANIRUDH-VADERA
- LeetCode for algorithmic consistency and contest history: leetcode.com/u/anirudh00711
I like problems that require equal parts engineering depth, business context, and operational patience. If the system has money movement, scale, reconciliation, and a pager attached to it, I am probably interested.