Amrit Sena

Backend engineer with 3 years at Autodesk building and operating high-throughput, cloud-native services on AWS. Scaled event-ingestion pipelines 12x, led cloud migrations and CI/CD modernisation, and own production reliability end to end: on-call, disaster recovery, and observability.

Software Development Engineer II at Autodesk (2025-04 – Present)

• Scaled the licence-misuse detection pipeline from 40 TPS to 500 TPS (12x) by redesigning Kinesis shard allocation and tuning ECS autoscaling, leading a team of 2 across infrastructure, load testing and observability; held p99 latency under 200 ms at peak.
• Built the bulk data-upload flow handling 1M+ records per run, adding idempotent retries and dead-letter handling for Kinesis shard failures and cutting failed-batch reprocessing by X percent.
• Built an AI agent that automates manual regression testing across desktop applications, replacing 64 hours of manual QA per release.
• Worked with PMs to built dshboard for tracking the revenue funnel for non-compliant to compliant conversion
• Delivered self-serve ingestion onto the internal data platform, enabling 200 teams to consume datasets at near real-time.
• Migrated 20 build and deploy pipelines from legacy Jenkins to CBCI and Spinnaker.
• Primary on-call for 20 production services: resolved P1 incidents across the geoinfo pipeline and products-service availability, and led DR exercises validating multi-region failover with an RTO of under 3 minutes.

Software Development Engineer I at Autodesk (2023-11 – 2025-03)

• Delivered the entitlement backend for Capacity-Based Licensing: a queue consumer polling SQS to reconcile seat usage asynchronously, plus a read API serving entitlement checks at sub-100ms p99 under 100 requests/second.
• Wrote the core processing Lambda for the event-reporting pipeline in Go, handling 50M events/day at sub-100ms p99 execution time while keeping concurrency.
• Developed a zero-downtime AuthEvent schema migration across licensing event producers and consumers improving data-quality for better analytics.
• Raised automated test coverage from 40 to 80 percent by building a Python BDD framework covering 5 tier-1 platform services.
• Migrated services onto the internal CloudOS platform (ECS, Lambda) and instrumented Dynatrace dashboards and alerting for 5 services.

B.Tech in Computer Science and Engineering – Institute of Technical Education and Research (2019-07 – 2023-06)