Biosonic

Overview

BioSonic is a bat-acoustic research platform used by 5,000+ scientists worldwide. I worked as a shadow developer building the platform end to end, shipping production-grade infrastructure handling data at a scale most platforms never encounter.

---

The Engineering Challenge

The core problem was scale. Bat acoustic research generates data in the millions of records, so large that even Supabase's newly launched analytics bucket couldn't handle it at the time. Custom solutions were engineered from scratch to ingest, store, process, and display this data reliably on the web.

---

What I Built

• Browser-Based Batch Upload System, Processed millions of files in a single session using the WHATWG File System API, with no Electron dependency
• 1,000-File Batch Processing, Chunked upload architecture for fault-tolerant large-scale data handling
• Custom RPC Functions, Built Supabase RPC calls for large data aggregations and complex data handling that standard queries couldn't efficiently support
• Custom SQL Layer, Highly optimized queries on Supabase and PostgreSQL handling millions of records where off-the-shelf solutions fell short
• Web Data Visualization, Rendered massive datasets in the browser without performance degradation
• Real-Time Processing, Audio files processed and surfaced to researchers as they arrive

---

The Scale Problem

When Supabase launched their analytics bucket, our dataset was already beyond what it could handle. The solution required custom RPC functions, query optimization, batching strategies, and architecture decisions built specifically for this level of scale.

---

Tech Stack

Built with React, TypeScript, Supabase, custom SQL, and Supabase RPC, deployed for production use by thousands of active researchers globally.