Hi, I'm Anas Ikhlaq

Point-of-sale desktop application with real-time inventory tracking and transaction handling.

Internal tool for managing employee records, roles, and scheduling across teams.

Interactive quiz platform with live scoring, timed questions, and result breakdowns.

Automated workflow that compiles and delivers a daily weather digest via email.

Automated pipeline that aggregates news and pushes a digest straight to Telegram.

Full-stack e-commerce platform for a leather goods store, with product catalog and checkout flow.

Developed an intelligent FAQ chatbot that integrates Google Gemini with a spreadsheet-based knowledge base to deliver real-time responses. Designed an automated escalation process that collects user contact information and logs unresolved queries into Google Sheets for support team review.

Developed and executed side-by-side evaluations of AI coding assistants by simulating real-world software development scenarios. Assessed model performance based on code quality, usability, accuracy, and practical effectiveness in development workflows.

Evaluated AI-generated code for correctness, executability, robustness, and overall quality using structured assessment frameworks. Analyzed model outputs and validation results to ensure generated solutions met functional and performance requirements.

Created high-quality reference implementations and comprehensive unit tests for RLHF datasets. Validated code functionality through positive, negative, and edge-case testing in clean execution environments.

Reviewed software engineering issues from open-source GitHub repositories by analyzing bug reports, code fixes, and associated test cases. Verified that test modifications accurately validated implemented solutions and maintained software reliability.

Developed a high-performance e-commerce storefront for motorcycle components using Next.js 15+, TypeScript, and Tailwind CSS v4, integrated with Supabase for secure, server-side data processing and Google OAuth. Engineered a tamper-proof checkout pipeline that recalculates real-time pricing server-side to prevent client-side exploitation, paired with Make.com webhook automation for real-time customer order confirmations.

Developed a Retrieval-Augmented Generation (RAG) conversational agent on n8n to serve as an Islamic knowledge assistant. Configured an automated pipeline that ingests source documentation from Google Drive, stores and processes vector embeddings within Supabase, and utilizes the Google Gemini Chat Model to generate context-aware, accurately sourced answers to user inquiries.