Muhammad A.

I’m a Data Scientist (Master’s in Data Science) with 5+ years of experience building production-grade data and AI solutions, especially LLM/Agentic AI systems, evaluation automation, and end-to-end deployments. My work spans NLP/Generative AI (RLHF, RAG, model evaluation) and scalable data platforms, with a track record of improving reliability, speed, and outcomes across real datasets. Recently, I’ve been working on AI Agents and real-API scenario engineering, where I helped operationalize dataset porting to real systems (e.g., Slack/Gmail/Calendar-style APIs and commerce/expense platforms) and built automation to generate high-quality notebook workflows from API documentation, reducing per-scenario build time from 120 minutes to 45 minutes (2.7× throughput). I also bring hands-on experience with multi-agent workflows and voice AI solutions (AI voice assistants handling customer calls), including real-time transcription and summarization pipelines. Core strengths: - Agentic AI + LLM Systems: RLHF, RAG, prompt engineering, model evaluation, automation frameworks - Speech + Vision: Whisper-based real-time transcription; MediaPipe-based assessment systems - Data Engineering + BI: AWS Redshift, DOMO/Power BI, real-time analytics at scale (100M+ rows) - Web Scraping & ETL: robust Selenium pipelines (e.g., 12M+ records ingested in 48 hours) Services I offer: - Agentic AI systems (multi-agent workflows, tool integrations) + LLM apps (RAG chatbots, evaluation/autoraters) - Data platforms, pipelines, dashboards, and migrations (Redshift, DOMO, Power BI) - API development & deployment (Flask, Azure Web App Service) - Data cleaning, feature engineering, and ML modeling (measurable lifts, production mindset) If you want a partner who can translate your goal into a clear implementation plan and deliver reliable results, I’m happy to discuss your project and propose the quickest path to impact.