- Master’s student (M.S. Computer Science, Machine Learning) @ George Mason University (GPA 3.8/4.0)
- Software Engineer with full-stack chops and a passion for applied AI/ML, especially:
- Retrieval-Augmented Generation (RAG) and LLM ops
- Zero-shot anomaly detection & multimodal models
- Front-end craftsmanship with React + Tailwind
- I love turning research ideas into production-ready code, writing clean APIs, and automating everything in between.
- Building smarter RAG pipelines (HyDE, multi-query, eval agents)
- Optimizing model inference for edge & cloud (quantization / pruning)
- Contributing to open-source CSS/React component libraries
| Category | Tech |
|---|---|
| Languages | Python · C/C++ · JavaScript/TypeScript · SQL |
| ML / DL | PyTorch · TensorFlow · Hugging Face · LangChain |
| Web | React · Redux · Tailwind CSS · Spring Boot · REST/GraphQL |
| Data / Infra | PostgreSQL · MongoDB · Docker · AWS · GitHub Actions |
| Testing / QA | Selenium · PyTest · JUnit |
| Project | What it does | Key Tech |
|---|---|---|
| SentinelAI | Natural-language → auto-generated analytics dashboards | OpenAI API · LangChain · React |
| AnomalyCLIP | Zero-shot visual anomaly detection with textual prompt tuning (AUROC 94.7 +) | CLIP · DPAM · PyTorch |
| Ray-RAG | <100 ms question-answering on Ray docs, +24 % accuracy vs. vanilla | Python · LangChain · PostgreSQL |
| Text-to-SQL T5 | Fine-tuned & prompted T5 achieving 0.627 F1 on domain data | PyTorch · HF Transformers |
- 💬 Ask me about ML engineering, RAG workflows, or React UI/UX
- 📧 Reach me: skakarl3@gmu.edu
- 🌐 More about me on LinkedIn
“Code is craft; models are stories; great products combine the two.” — me, probably