Sumit Sharma

Software Engineer

Seattle, Washington, United States6 yrs 2 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Proven track record in backend architecture and distributed systems.
Led complex projects that significantly improved system performance.
Strong experience in mentoring and cross-functional collaboration.

Stackforce AI infers this person is a Backend and Infrastructure Engineer specializing in high-performance systems for B2C applications.

Contact

Skills

Core Skills

Distributed SystemsBackend ArchitectureAi IntegrationSystems DesignReal-time ProcessingBackend DevelopmentMicroservicesData EngineeringUser Experience DesignInfrastructure ManagementQuality Assurance

Other Skills

AIAPI DevelopmentAPI OptimizationAWSAWS EMRAWS GlueAWS LambdaAmazon DynamodbAmazon Elastic MapReduce (EMR)Amazon S3Amazon Web Services (AWS)Apache AirflowApache FlinkApache KafkaApache Pinot

About

I'm a Backend and Infrastructure Engineer based in Seattle (US work authorisation secured via L2S EAD). I'm passionate about building the resilient, high-performance systems that power massive-scale consumer experiences, and I'm eager to bring my experience to a new team here in the Pacific Northwest. My 5+ years across Uber, Swiggy, and Amazon have been focused on delivering mission-critical features using distributed systems. I specialise in creating reliable backend architectures, optimising real-time data flows, and implementing features that leverage AI to boost user engagement and monetisation. I believe great engineering is about more than just clean code - it's about solving the right problems. I have a proven track record of leading complex projects, modernising older infrastructure, and working cross-functionally with Product and Data Science. You’ll often find me mentoring engineers, leading technical discussions, and simplifying processes to improve overall team velocity.

Experience

6 yrs 2 mos

Total Experience

2 yrs

Average Tenure

2 yrs

Current Experience

Uber

Software Engineer II

Jun 2024 – Present · 2 yrs · Seattle, Washington, United States · On-site

Re-architected the Uber Eats Manager Stores page from a client-heavy flow to backend-driven search/pagination on OpenSearch with Kafka+Flink (permissions embedded in the index), cutting p99 from 2–3s to 100ms and client memory by 90% for merchants managing 15K stores; technically delivered a hybrid batch/stream pipeline with multi-layer idempotent dedupe and safeguards for out-of-order sales/order updates.
Led end-to-end menu content optimization by shipping an image-scraping pipeline and layering AI-powered image suggestions plus LLM-generated descriptions with online evaluation guardrails; boosted checkout conversions by 18% and drove an average 23% revenue lift, with the hard parts being reliable pipelines, safe low-latency orchestration, and quality controls at scale.
Redesigned buffer hours so merchants can set last-order windows in precise 1–5 minute increments (vs. 15-minute buckets), reducing overestimated buffers and extending availability to drive an estimated $26M annual GB uplift globally; required platform-level reindexing across SIA ingestion/serving and a runtime filter to keep unorderable stores out of search/feed while controlling index size and latency.
Built a real-time notification system that consumes store-status events from Kafka in Flink to send email/SMS to admins about downtime/pauses with noise-suppressing guardrails (dedupe, rate limits, one-per-24h), then enhanced it with potential-sales-lost enrichment and a secure one-click unpause link (hashed/TTL tokens, pessimistic-locking thresholds), improving unpause speed and online rates; technically centered on backpressure-safe streaming, idempotent delivery, and end-to-end observability

Apache PinotApache FlinkPrestoApache SparkHadoopApache Kafka+9

Swiggy

Software Developer II

Mar 2022 – Apr 2024 · 2 yrs 1 mo · Bengaluru, Karnataka, India · Remote

Built a scalable ingestion pipeline syncing customer chat data from external platforms into Swiggy’s data lake, saving $60K/month by eliminating reliance on third-party tools.
Developed a dynamic context generation pipeline for LLM-powered customer care bots, significantly reducing live agent ticket queue load by enabling the bot to handle ad-hoc queries.
Designed and implemented Pay-On-Delivery on Track Screen, enabling customers to complete payment after placing an order in case of payment failures. This feature eliminated 5% of order drops caused by failed payments.
Integrated iOS Live Activity tracking, reducing order-tracking API calls by 15% and improving customer experience with real-time updates.
Leading tech simplification across domains. Optimized 3+ services, reduced Deployment Units and
associated infrastructure from 15 to 6, eliminated 1,000+ outdated configurations, and achieved $6K+/month in cost savings. Formulated a plan to reduce infrastructure components by an additional 50%

Amazon Web Services (AWS)Spring FrameworkTechnical RequirementsMySQLMultithreadingDesign Patterns+15

Amazon

2 roles

Software Developer I

Aug 2020 – Mar 2022 · 1 yr 7 mos

Built a centralized schema service integrated into Ads ETL ingestion jobs, enabling runtime schema resolution and versioned deserialization. Eliminated schema drift-induced ingestion failures and reduced on-call incidents by 60%.
Built a workflow manager for AWS EMR jobs using Lambda + SQS, cutting resolution time from 2 hours to 10 minutes.
Developed core backend APIs for Amazon's internal Ads Data Platform, building highly scalable and low-latency solutions. My work enabled internal teams to centralize, query, and visualize business-critical data, improving data accessibility and analytical capabilities.

Amazon S3Spring FrameworkDistributed ComputingTechnical RequirementsElasticsearchMySQL+12

Intern

Jan 2020 – Jul 2020 · 6 mos

Automated the Amazon Ads testing lifecycle by engineering an integration framework that reduced the QA cycle time from 2 days to 1 minute and enabled continuous regression detection via CI/CD integration.

Amazon Web Services (AWS)REST APIsRemote Procedure Call (RPC)Spring FrameworkTechnical RequirementsCode Review+4