Working at the intersection of production ML infrastructure and high-scale backend engineering
| Domain | Impact |
|---|---|
| π§ ML Inference | gRPC layer on TF Serving β β throughput, β latency for nonlinear DL models |
| ποΈ Feature Platform | Varys Feature Store @ Flipkart β real-time ranking at national scale, multi-DC |
| π Data Pipelines | Spark + Azkaban β Aerospike; ranking features scaled 8 β 40 |
| π ML Observability | Druid cluster β billions of rows of feature telemetry, DS dashboards |
| ποΈ Scale Readiness | BBD / Opus NFRs, chaos playbooks β zero P0/P1 on peak load |
| π€ NLP Research | Hybrid extractiveβabstractive summarization pipeline @ Samsung PRISM |
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β β
β π Flipkart SDE-II Aug 2025 β Present β
β π Flipkart SDE-I Sep 2023 β Aug 2025 β
β π Byju's MTS Intern Sep 2022 β Sep 2023 β
β π¬ Samsung PRISM Research Intern Oct 2021 β Aug 2022 β
β β
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ


