ML Research Engineer · NLP & Agentic Systems · CS @ Arizona State University
- 🔬 ML Research/Engineering @ Pinpoint Technologies LLC (May 2026 – Present) — designing graph neural network models with PyTorch Geometric for richer relational representations on network data, plus data/feature pipelines and applied research on graph-based learning
- 🧠 Previously NLP Engineer Intern @ Time Pilot (Sep 2024 – Mar 2025) — FastAPI services connecting ML models to production apps, embedding models, prompt engineering, and fine-tuning multitask classification/generation models
- 🎓 B.S. Computer Science, Arizona State University (2025 – 2026) — GPA 3.56; coursework in knowledge representation, NLP, statistical machine learning, and large-scale data processing
- 🌱 Currently exploring continual learning for LLM agents and agentic systems design
- 📫 Reach me: dkumar70@asu.edu
Experience Engine — a continual-learning research prototype for LLM agents. Turns an agent's past trajectories into validated, reusable competence (record → evaluate → diagnose → induce → validate on held-out → update policy) instead of just retrievable memory, benchmarked against a stateless and a memory-only baseline on execution-graded QA tasks.
- Royce Carbowitz, Dheeraj Kumar. SPOQ: Specialist Orchestrated Queuing for Multi-Agent Software Engineering. arXiv:2606.03115
- Dheeraj Kumar, Vibhanshu Rana. Ontology Engineering for Trustworthy AI: A Production-Grounded Survey of Knowledge Graphs in the LLM Era. DOI:10.13140/RG.2.2.29422.16965
- Dheeraj Kumar. GCAF-Net: Gated Cross-Attention Fusion for Efficient Robotic Vision-Language Reasoning. DOI:10.13140/RG.2.2.12906.38089
- AI Scholar Program, Arizona State University
- Cintana Education Scholarship Program
- Devlab's Devhacks Season 2 — Category Winner
- 1st Place, Dutch Bros × ASU AI Broista Challenge 2025 (Dutch Bros Tempe HQ)
- 1st Place, AI Scholar Principled AI Hackathon
- Prime Minister's Scholarship — full undergraduate tuition waiver
