
Nikhil Saravade
Backend Engineer | Java, Spring Boot, Distributed Systems | AI/ML Systems Engineering
Talent
PuneMember since September 23, 2026
JavaPythonSQLGoSpring BootMicroservicesFastAPIPostgreSQLMySQLOracleJPADistributed systemsKubernetesOOPLLM inference internalsRoPEPaged KV-cacheAgentic tool-callingCross-encoder reranking
·
Hire this person
Send a job offer directly to this candidate
JavaPythonSQLGoSpring Boot
About
Backend Engineer with 2+ years building production-grade financial systems on Java, Spring Boot, and distributed data pipelines, with recent hands-on depth in AI/ML systems engineering, including from-scratch LLM inference internals, agentic tool-calling systems, retrieval-augmented generation, and model fine-tuning, each validated with real, statistically grounded evaluations rather than assumed results.
Experience
Systems Engineer (Backend Developer) at Tata Consultancy Services (2024-08 – Present)
- Architected and built a hedge-fund portfolio advisory platform (system of record for portfolio construction, rebalancing, and performance analytics) from the ground up, modeling a 150+ entity domain with CRUD, versioning, allocations, and NAV tracking over JPA repositories.
- Delivered a REST API of 130+ endpoints across 9 controllers with typed DTOs, Bean Validation, and OpenAPI-ready contracts, plus an async report-generation pipeline (@Async + job-status polling) that moved long-running analytics off HTTP threads for concurrent report processing.
- Engineered the quantitative core, including a rebalancing engine (cash injection/redemption, pending-transaction reconciliation, position-liquidity calculations), an analytics/optimization module (efficient-frontier, rolling-style, regression/statistics) fronting a MATLAB engine via a clean Java interface, and a configurable reporting engine (Apache POI + JFreeChart) generating 20+ report/table types.
- Automated 4 Spring Boot batch jobs handling end-to-end ingestion of custodial and fund-pricing feeds (fetch, decrypt password-protected ZIPs, parse 3 Excel formats via Apache POI, reconcile, and load), replacing a manual daily reconciliation process.
- Built a multi-threaded performance loader (ExecutorService thread pool) bulk-loading NAV/performance data across two databases (Oracle + MySQL) concurrently, reducing nightly load time by ~45%.
- Standardized deployment of all batch jobs as environment-aware Spring Boot jars across 4 environments (DEV/UAT/PROD/COB) with externalized config, log rotation, and email failure alerting.
Education
B.Tech in Information Technology – JSPM's Rajarshi Shahu College of Engineering (2020 – 2024)
Skills
Backend EngineerJavaSpring BootDistributed SystemsAI/ML Systems Engineering



