Mar 2025 – Oct 2025 · 7 mos · Bengaluru, Karnataka, India · Hybrid

• Team : Plan and Discover , Core Logistics (Last Mile , Flipkart)
• ● Engineered and launched key logistics features, including Past Delivery Based Allocation and Direct Asset Bagging
• , directly resulting in a combined operational cost saving of over ₹3.5 Crore through reduced misroutes and
• manual sortation.
• ● Drastically optimized a critical P0 API by slashing P99 latency by over 90% (from ~30s to under 600ms) by
• implementing a high-performance Caffeine in-memory cache and replacing inefficient data access patterns, which
• eliminated critical timeouts.
• ● Pioneered and built a reusable, end-to-end data pipeline (TiDB → TiCDC → Pulsar → Spark) to stream data into the Flipkart
• Data Platform, establishing a new team-wide standard and empowering business with crucial misroute analytics.
• ● Enhanced system scalability and reliability by leading the migration of core services from MySQL to TiDB to handle data
• explosion, and implemented an Aerospike-based distributed lock to eliminate critical race conditions in a multi-pod
• environment.
• ● Designed a multi-tenant hub allocation system to enable flexible, tenant-specific logistics strategies, significantly
• improving last-mile network efficiency and operational control.
• ● Designed a shadow pipeline using Kafka queue for risk-free, real-time A/B testing of new allocation algorithms
• against live production traffic, enabling data-driven validation before deployment.
• ● Enhanced observability during peak sales events ('Big Billion Days') by creating comprehensive Grafana dashboards for
• all critical services.

Jul 2024 – Mar 2025 · 8 mos · Bangalore Urban, Karnataka, India · Hybrid

• Team - logistics and offline conversion
• ● Developed a city-based authorization for PF Approval Notification system, enhancing transaction-level control and
• boosting revenue by 20% through optimized PF charges and scheme flexibility.
• ● Implemented PAN-Aadhaar validation at the top of the funnel, ensuring compliant loan applications and reducing
• downstream processing delays.
• ● Introduced channel-based lead attribution rules to accept or reject leads based on their initial transaction source, saving
• 30% in commission expenses by preventing payouts for leads sourced by internal agents.
• ● Implemented a PoC using Nginx to embed an Appsmith page in a public Bubble app while keeping Appsmith behind a VPN, enabling secure access without needing a separate pod , reducing server costs.
• ● Designed a hierarchical lead management platform with tree-structured visibility for agent-manager metrics using
• MongoDB aggregation and graph lookup for efficient data access.
• ● Streamlined user metrics updates with Redis to reduce database writes and implemented cron jobs for Redis-database
• sync and DFS-based hierarchical updates.
• ● Automated agent onboarding, role updates, and archiving via a JavaMail API email listener for HR CSV processing, and
• enabled granular permissions with PBAC using Spring filters and PreAuthority.
• ● Developed a check-in feature && enabled geo-tagging verification to improve agent activity monitoring by managers .
• ● Built a lead metrics dashboard using MongoDB Aggregation Pipeline for actionable analytics on lead disposition and
• agent activities.
• ● Enabled Coborrower loan creation across the conversion , scheduling and appraisal flow which lead to a major revenue boost .
• ● Integrated AA RECURRING flow which helped in better analysis of customers financials && expenditures .

Jan 2024 – Jul 2024 · 6 mos · Bangalore Urban, Karnataka, India · Hybrid

• Team - Logistics and Offline conversion
• Enhanced the Sales App’s PC digitization ow with PDF uploads, advanced search filters, and streamlined update
• options, reducing team dependency, boosting efciency, and minimizing manual tasks.
• Designed and implemented a new Backend for Sales App’s schedule flow which included PII data migration , made the
• system RBI compliant and saving upto 200+ dollars per month by reducing workow units .
• Optimized Pricing APIs by implementing caching , improving database queries, and removing redundant calls, reducing
• latency and enhancing data retrieval performance by 30% .
• Engaged in design discussions and code reviews, utilized CI/CD pipelines for deployment, resolved critical on-call issues,
• and leveraged Grafana and New Relic for monitoring