Achu Mukundan
Fullstack / Backend Engineer
Building local-first AI tools, developer systems, and production-grade infrastructure.
Experience
Junior Developer
May 2023 — May 2024Replicant AI
Worked on production conversational AI systems, internal LLM tooling, and NLP testing infrastructure for a platform serving 10K+ daily users.
- Built and shipped production conversational AI bots in TypeScript/Node.js serving 10K+ daily users
- Implemented low-latency conversation-state handlers processing 500K+ bot events/day with approximately 150ms median latency
- Led development of an internal LLM chat interface used by 200+ company employees
- Migrated legacy workflows to LLM-based data collection, improving bot/data accuracy by approximately 30%
- Designed timezone-aware scheduling/detection modules that reduced bot misfires by approximately 15% and improved CSAT
- Built an automated NLP intent-recognition testing suite with Jest, reaching 90%+ test coverage and reducing regressions
- Introduced structured logging and health endpoints, improving debugging under elevated load
Education
University of Toronto
Graduated 2025BASc, Computer Engineering
Relevant coursework: Algorithms and Data Structures, Machine Learning, Software Engineering, Databases, Computer Networks, Computer Security, Operating Systems.
Featured Projects
A lightweight compatibility gateway for running local models through OpenAI- and Anthropic-style APIs.
An experimental terminal coding agent built around local inference, OpenAI-compatible model gateways, and small/quantized model workflows. Currently in closed alpha.
An electronic music world spanning releases, events, merch, visual systems, and community experiments.
A local-first creative operating system with SQLite FTS5 search, local LLM summarization, and structured retrieval workflows.
Docker-based distributed storage with Cloudflare Zero Trust access, observability, and concurrent metadata pipelines.