Iremide Adeyanju

I’m a Backend Software Engineer focused on shipping AI-integrated backend platforms using Python and modern API/service patterns. I care about latency, reliability, and building systems that don’t break when upstream dependencies degrade.

Most recently, at FindProducts Inc., I cut product-search latency by 50% and delivered ~30x query speedups by integrating Elasticsearch with PostgreSQL and tuning analyzers for multi-term vendor queries across a multi-million-row catalog. I also increased API throughput ~3x by moving scraping and fan-out work into Celery + Redis workers, keeping request workers available for latency-sensitive reads.

I’ve built production-oriented backends with DRF + Supabase, including JWT auth, idempotent write paths, and webhook-driven cache invalidation. On the API side, I reduced list-endpoint payloads ~60% using DRF field-selection serializers and cursor pagination, improving p95 response for mobile clients.

On the AI systems side, I designed a multi-provider LLM gateway with rate limiting and per-provider circuit breakers, plus caching that reduced duplicate outbound LLM calls ~40% in benchmarks. In hackathon projects, I’ve delivered voice and multilingual pipelines (Gemini + Deepgram) and RAG-driven experiences that improve retrieval and conversational performance over large document sets.