Arsh Ergon

Architected microservices and GraphQL APIs serving 800+ users at 93.9% uptime and built FeeFlow school management platform for 500+ students automating 80% of billing; delivered SvelteKit edge caching that reduced load times to sub-1s and drove 80% organic traffic growth.

Optimized Llama-2 fine-tuning with QLoRA reducing memory overhead by 50% and improved LLM response accuracy by 20% via vector embeddings and TDD-driven inference pipelines during a US-remote contract.

Built cross-framework primitives for an Ivy ML transpiler enabling TensorFlow/PyTorch/JAX interoperability and delivered a 10x performance gain in tensor sorting through low-level optimizations.

Delivered 50+ client contracts including Django/PostgreSQL school management systems for 500+ users and integrated Twilio APIs to reduce manual admin work by 80%.

Master of Computer Applications with focus on data structures, system design, distributed systems patterns, and advanced Rust programming; prepared for ML engineering roles.

Focused study program in data structures, algorithms, and system design fundamentals to prepare for technical interviews and ML engineering roles.

Bachelor of Technology in Information Technology completed concurrently with freelance development, delivering production systems and serving 50+ global clients.