Daryl Micah
Bengaluru
Full Stack Developer
Full Stack Engineer with 2.5+ years designing and building services end-to-end, schema through API through internal UI, in Python (FastAPI), PostgreSQL, and React/TypeScript. Real experience with idempotent data pipelines, authenticated machine-to-machine integrations, and query-level performance work. Uses Claude Code daily as a normal part of shipping, and owns everything that lands regardless of what drafted it.
Selected work
3 piecesCCOS – Creator Campaign Operating System
Full-stack CRM built around a derived-metrics engine (computed, not duplicated state) with Redis-cached, queue-driven async data collectors backed by a normalized PostgreSQL schema.
- My contribution
- Full-stack CRM built around a derived-metrics engine (computed, not duplicated state) with Redis-cached, queue-driven async data collectors backed by a normalized PostgreSQL schema.
Avert Agent
Automation tool that detects external API/SDK breaking changes via incremental code indexing, with a typed data model and test coverage (pytest, Vitest) across the pipeline.
Work experience
– Present
Full Stack Developer
Dom Ventas (Broadwing Labs) · Bengaluru, KA
* Built a serverless, queue-based data pipeline on AWS (Lambda/SQS/S3) integrating the authenticated Amazon SP-API, normalizing 8+ data entities into PostgreSQL with idempotent, replayable backfills. * Delivered a FastAPI + React (Next.js) ecommerce dashboard end-to-end, schema through API through UI, with JWT-based authentication and streaming CSV/XLSX upload workflows, spanning ~400 commits across 4 services. * Uses Claude Code daily as a normal part of the development workflow, building internal automation scripts and one-off dashboards, and reviews everything it produces before it ships.
–
Software Developer
Logicsoft International · Gurugram, Haryana
* Drove a 70% cut in end-to-end API latency across multi-tenant Node.js/PostgreSQL services through query profiling, composite indexing, and eliminating N+1 patterns. * Designed JWT-based authentication with RBAC across a distributed, multi-tenant backend, scaling from single-tenant to high-concurrency multi-tenant with zero downtime. * Reviewed pull requests and enforced test coverage standards across the team, catching regressions before deployment.
–
Technology Intern
EigenGram · Remote
* Improved LLM and vision-model inference throughput by 25% through model quantization and structured pruning, enabling the same workload on lower-cost infrastructure. * Built FastAPI backend services for real-time LLM/vision inference pipelines, cutting service latency 15% via optimized async request handling.
–
Software Engineer Intern
Volkswagen IT Services · Pune, Maharashtra
* Improved data retrieval speed by 50% on an enterprise security platform by redesigning REST API response structures and React data-fetching strategies.