Maya Sengupta Tech Lead
San Francisco, CA • maya.sengupta@gmail.com • +1 415-555-0163
Profile Summary
- Tech Lead with 8 years of experience building consumer marketplace products across checkout, payments and booking flows, leading a 7-engineer team as its technical owner while still shipping code every week.
- Hands-on with languages (Java, TypeScript), frameworks (Spring Boot, React), cloud (AWS, Kubernetes), data (PostgreSQL, Apache Kafka) and monitoring (Datadog).
- Deep expertise in technical direction for one team, design docs and design reviews, breaking projects into shippable milestones and service reliability, using staged rollouts and blameless postmortems to ship features the team can run on its own after launch.
- Works every day with the engineering manager, the product manager, design and 2 partner teams in sprint planning, design reviews and quarterly roadmap sessions, turning roadmap goals into tickets the team can pick up.
- Mentors engineers through weekly pairing and design doc feedback, wrote the team’s code review checklist, and runs the weekly on-call handoff and the monthly postmortem review.
Technical Skills
- Languages & Frameworks:
- Java, TypeScript, Spring Boot, React, Kotlin, Node.js, Python
- Cloud & Infrastructure:
- AWS, Kubernetes, Docker, Terraform, Helm
- Databases & Messaging:
- PostgreSQL, Apache Kafka, Redis, Amazon DynamoDB, Elasticsearch
- Architecture & Design:
- System design, design docs, RFCs, ADRs, microservices, event-driven architecture, REST APIs, gRPC
- Testing & CI/CD:
- Code review, CI/CD (GitHub Actions), JUnit, Jest, integration testing, feature flags (LaunchDarkly), trunk-based development
- Observability & On-Call:
- Datadog, Grafana, OpenTelemetry, PagerDuty, SLOs, error budgets, postmortems, runbooks
- Planning & Team Practices:
- Sprint planning, Jira, estimation, pair programming, mentoring, technical interviewing, Confluence
Education
Work Experience
- Set the technical direction for the 7-engineer Guest Checkout team in a yearly tech plan agreed with the EM and PM, moving 23 checkout endpoints off the Ruby on Rails monolith into Java services while keeping 20% of every sprint for tech debt.
- Wrote the design doc for split payments on group bookings and ran its design review with the payments, risk and security teams, choosing an idempotent payment API over client-side retries, with 0 double charges across the first 1.2M bookings.
- Kept writing production code every week, including the price quote cache in Java on Redis, which cut checkout page p95 load time from 1.4 s to 620 ms on web and mobile.
- Broke the monolith migration into 9 milestones of 2 to 3-day Jira tickets, tracked dependencies on the payments and pricing teams in a weekly sync, and cut tickets blocked for more than 3 days from 18% to 4%.
- Wrote the team’s code review checklist with a 1-business-day turnaround target and added contract tests and coverage gates to GitHub Actions CI, cutting checkout bugs that reached production from 11 a quarter to 3.
- Planned each quarter with the engineering manager and the product manager, traded scope for dates in a shared roadmap doc, and shipped releases behind LaunchDarkly flags with staged rollouts, landing 10 of 11 committed launches on time over 2 years.
- Mentored 4 engineers through weekly pairing and design doc reviews and wrote input for their promotion cases, with 3 promoted to senior engineer, and joined the interview loop for every hire on the team.
- Owned on-call health for 6 checkout services, tied alerts to SLOs in Datadog, wrote runbooks and ran blameless postmortems, cutting weekly on-call pages from 14 to 3.
- Rebuilt the nightly tax-loss harvesting job as a Java service reading account events from Apache Kafka, cutting the run from 5 hours to 70 minutes.
- Acted as tech lead for 3 engineers on the cash account transfer service, writing its first design doc and shipping it in 10 weeks, 2 ahead of the planned date.
- Added contract tests between the iOS, Android and web clients and the transfers API, catching 27 breaking changes before release in the first year.