Abdoul Rachid Ouédraogo (韋正)

Applied AI Researcher @ Taiwan Polar Institute · Founder of POUK YAM

---

Building agentic AI platforms for national agencies: MCP servers, tool-calling analysts, data collectors, forecasting models, dashboards.

Background: 6+ years across R&D, application development, and research, including LLM agents with RAG and ML shipped in a commercial product.

---

🚀 Featured projects

AquaScope ⭐ 28 Open-source Python toolkit that unifies global water data under one API, then puts an agent on top of it. 29 collectors behind one Pydantic schema: USGS, UK EA, Hub'Eau, GRDC, FAO, GEMStat, EU WFD MCP server — Claude or Cursor gets stations, series, and cited flood frequency as real tools Studio: a six-role agent crew (scout, methodologist, analysts, critic, author) plans, runs, and returns a Word + Excel + notebook bundle HydroGym: gym-style RL environment for model calibration on real basins 1,000+ tests · CAMELS-validated · 24 outside contributors · Zenodo-archived	Agronaut Multimodal tool-calling agent over a deterministic, cited engineering core. Zero proprietary APIs. LLM collects and routes; a validation gate guards a pure, tested trust zone that does the math Photos and voice notes on Telegram, WhatsApp, and the web — a guard strips any fabricated reading, and a cited table returns a ranked differential, never a verdict RAG measured, not asserted: hit 0.879 · recall 0.833 · MAP 0.604 · 8/10 off-topic queries refused. Nine techniques benchmarked, four ship, every verdict recorded Faithfulness + citation-accuracy eval; per-turn tracing that records shape, never content 1,235 tests · advice-safety golden set enforced in CI · DPG compliance pack
HydroPhysicsAI A differentiable digital twin of Taiwan's Choushui alluvial fan: pumping policy in, groundwater heads and land subsidence out, gated against held-out data at every stage. 174 monitoring wells, 116,769 metered pumps, 798 leveling benchmarks, 2012–2022, on a 1 km four-aquifer grid Adjoint-calibrated flow solver with a physical pumping stress: held-out-well R² 0.804 vs 0.702 for inverse-distance interpolation Compaction column calibrated on the leveling network: R² 0.546 out of fold; full chain hindcast 0.526 on 798 sites Policy projections to 2032 with a fold and posterior ensemble; 18 members × 3 policies in 609 s on one GPU, PhysicsNeMo FNO surrogate 13–34× faster again Every gate verdict recorded, including the failed configurations and the held-out-years result	Mooré-Voice Translation and speech recognition for Mooré, a language of ~8M people and almost no training data. Curated 205,271 parallel pairs across 4 directions: LID-gated, fragment-filtered, decontaminated against FLORES-200 LoRA fine-tunes of NLLB-200 600M and 3.3B on a single consumer RTX 4070 ASR corpus of 37,654 utterances / 85 h; Whisper-small at 34.1% WER, benchmarked against an MMS-1b zero-shot baseline Three adapters published on the Hub · BLEU / chrF++ on FLORES-200 devtest · Colab demo
paper-agent ⭐ 8 · A Claude Code plugin that turns the agent into a disciplined manuscript collaborator — five strict modes (draft, review, revise, proofread, audit), Semantic Scholar MCP citation resolution with no API key, anti-fabrication guardrails, and a clean .docx round-trip. Ships journal profiles for Hydrogeology Journal, JHRS, and IEEE TIM, plus a generic quantitative-science profile. Install with one command: /plugin marketplace add Rekin226/paper-agent.

---

🛠️ Tech stack

PyTorch · Hugging Face (transformers, PEFT, datasets) · LangChain · LangGraph · AutoGen · MCP · CUDA / bf16 · NVIDIA PhysicsNeMo · RAG & vector search (FAISS) · Ollama / vLLM · pandas · NumPy · SciPy · FastAPI · Django · PostgreSQL · Docker · GitHub Actions · pytest · Streamlit · Gradio

🔬 Interests

Agentic AI & tool use · LLM fine-tuning for low-resource languages · Evaluation, guardrails & AI safety in production · Physics-informed & scientific ML · Uncertainty quantification · AI as a digital public good

---

📈 GitHub activity

---

Open to conversations about AI/ML engineering roles and collaborations. LinkedIn · rachidouedraogo.com

_{Building AI for science · Taichung, Taiwan}