Director of Software Engineering
Director of Software Engineering
- Joined as a founding member of a newly assembled engineering leadership team — alongside Directors of DevOps, QA, and Product — to transform a fragmented organization formed through rapid startup acquisitions into an enterprise-grade software development operation. Within 6 months, led a geographically distributed team of 100+ engineers across Mobile Development, Backend, Mapping SDK, and AI spanning the US, India, Mexico, and Eastern Europe, delivering an ecosystem of 6 products sustaining rapidannual growth — powered by cellular IoT infrastructure supporting millions of live cameras, billions of stored media files, and billions of new uploads annually.
- Architected a full engineering transformation from reactive firefighting to controlled, predictable delivery — establishing Amazon Leadership Principles as a cultural foundation, JIRA/Confluence as a single source of truth, Definition of Done enforcement, RACI frameworks, and structured escalation paths across a 100+ engineers organization spanning 4 geographies.
- Eliminated chronic misalignment between Software, Product, and Electrical Engineering teams by instituting weekly cross-functional syncs, bi-weekly product demos, and transparent roadmap visibility — replacing ad-hoc communication with structured, accountable processes.
- Implemented engineering analytics platform providing leadership with objective visibility into developer velocity, code quality, delivery throughput, and team health metrics across all 5 products — enabling data-driven performance management at scale.
- Inherited a stalled DB partitioning initiative with a failed September 2025 production attempt; diagnosed root causes, applied lessons learned, and drove full delivery of PostgreSQL table partitioning on an Aurora PostgreSQL database holding 4B+ records and growing exponentially — projected to reach 8B by end of 2026 — achieving ~20x faster device-based queries and ~10x improved user-based lookups, ensuring stability through peak hunting season.
- Sponsored and drove cross-functional delivery with DevOps to migrate media serving from S3 pre-signed URLs to a CloudFront CDN architecture — reducing API response payload by 87% (774 KB → 98 KB), cutting image URL size by 96%, improving gallery load times from 5s to ~1.5s on mobile, and achieving ~22x faster image delivery on cache hits; validated at 225,000 RPS.
- Identified and drove end-to-end redesign of the image colorization pipeline from bulk processing to on-demand, eliminating ~90% of wasted GPU compute and delivering ~$600K in annualized AWS savings — shipped via coordinated multi-service rollout with backend feature flags and mandatory mobile force-update.
- Led migration from Authorize.net to CyberSource to unblock EU market expansion and improve platform reliability — delivered with full staging validation and rollback procedures.
- Partnered with DevOps to complete DB right-sizing across all RDS instances, delivering $35,500/month ($420K annualized) in infrastructure savings with zero performance impact.
- Established clear ownership and cross-functional alignment between the AI and Engineering teams to accelerate YOLO (Ultralytics) deployment to production — replacing Detectron2 for real-time animal detection across 1.5M cameras, achieving 2.5x inference improvement and $7,340/month in compute savings.
- Delivered a production-grade AI support chatbot on AWS Bedrock (Claude Sonnet) with RAG pipeline, WebSocket streaming, and live device context injection — targeting a critical retention gap where poor customer support contributed to up to 40% of subscriber churn; system autonomously resolves 80% of support requests with seamless Zendesk escalation and full conversation context handover; compliance-mapped to CCPA, GDPR, and EU AI Act.
- Leading development of a Wildlife Intelligence Platform — ML-powered animal activity prediction system trained on 4B+ trail camera records using LightGBM on SageMaker, with automated MLOps pipeline and multi-factor inputs including weather, barometric pressure, moon phase, and GPS coordinates.
- Drove organization-wide AI tool adoption from near zero to active daily use across 97 engineers — rolling out Claude, Cursor, and Kiro across code development, code review, and test case creation, establishing AI-augmented engineering as standard practice.
- Deployed AI-powered code review agent across 3 products — identifying 443 findings and catching 80 defects before reaching QA or production, with 49-minute median feedback time and 68% PR coverage in the first month.
- Introduced Sentry for mobile observability and in-app bug reporting — within the first 3 months of deployment, identified and resolved 80 critical crash-causing defects across 3 mobile applications with 9M+ downloads, shifting defect detection from user-reported to pre-impact.
- Modernized Agile practices and CI/CD pipelines, reducing release cycles from 3 weeks to 2 weeks with on-demand hotfix capability; bugs down 23% and stories delivered up 26% quarter-over-quarter.
- Defined and executed multi-horizon technical roadmaps across 5 products in the first 6 months, partnering with C-level and product leadership to align engineering investment with business priorities and EU market expansion.
- Mentored engineers and emerging leaders across 4 geographies, embedding a culture of ownership, quality accountability, blameless retrospectives, and continuous learning.
