Open to opportunities

Jerrick Godwin

@jerrickgodwin

Message

I am a software engineer specializing in full-stack and LLM systems.

Sri Lanka

Message

What I'm looking for

I seek a collaborative, growth-focused engineering role building production LLM and full-stack systems, with opportunities to mentor teammates, work in agile environments, and solve ML infra and product challenges end-to-end.

I am a software engineer with over a year of experience building full‑stack applications and production LLM systems at CodeGen, where I implemented AI agents, hybrid RAG architectures, and optimized LLM fine‑tuning workflows. I have improved developer efficiency through tools that analyze commits and AST-based indexing, and reduced manual QA by 90% via automated test suites. I co-authored a 2025 conference paper on integrating domain-specific knowledge into LLMs and hold a BSc (Hons) in Computer Science.

I focus on pragmatic problem solving, hyperparameter optimization, and low‑memory, high‑performance ML deployments using techniques like QLoRA, LoRA, vLLM, and FlashAttention. I seek to contribute to agile teams where I can mentor peers, build impactful products, and scale AI-driven features in production.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer

Current

CodeGen Innovations (Pvt) Ltd

May 2025 - Present (1 year 1 month)

Implemented an AI agent to forecast purchase trends and events impacting customer interest for MCP. Improved LLM fine-tuning (LoRA, prompt engineering, vLLM, FlashAttention2), developed a hybrid RAG (dense+sparse, HNSW) for instant retrieval with low memory usage, and built a GitLab commit analysis tool to surface developer-efficiency insights.

Python LoRAs VLLM FlashAttention2 HNSW Prompt Engineering Gitlab

Associate Software Engineer

CodeGen International (Pvt) Ltd

Jul 2024 - Apr 2025 (9 months)

Developed a React-based web app with an advanced Canvas Editor for LLM interaction, significantly improving user experience across the company. Implemented a QA test suite with tool-augmented validation, reducing manual testing by 90%.

React TypeScript LLMs QA Automation Tool Augmented Validation Testing CI CD LLM

Software Engineering Intern

CodeGen International (Pvt) Ltd

Jul 2023 - Jul 2024 (1 year)

Fine-tuned LLMs with SFT and QLoRA achieving an 85% success rate on complex codebases and built an ETL pipeline that reduced manual effort by 60%. Implemented an AST-based codebase indexing algorithm to enable LLM fine-tuning on large repositories and improved developer efficiency by 15%.

SFT QLoRA ETL AST Based Indexing Fine Tuned LLMs Python PyTorch NLP