— Section 02 / About

A working notebook of applied AI.

AgereAI collects MVPs across LLM workflows, RAG, agent tooling and classical ML. Each project is published with rebuild notes, named test cases, and a status label that means what it says. The roadmap covers the full ML spectrum, from tabular baselines to LLM evaluation harnesses.

— 01 / Inside

What's in the portfolio

Working scope
  1. LLM automation workflows — e.g. Risk Analyzer: form → LLM → report → logs → email.
  2. RAG knowledge assistant with chunk-level citations (ingest + retrieval + constrained answers).
  3. Agent / tool-use orchestration with explicit human approval gates and audit logs.
  4. Core ML projects — tabular prediction with feature importance, forecasting, anomaly detection, NLP classification, computer vision.
  5. AI compliance / audit checklist grounded in recognized risk management thinking (NIST AI RMF).
  6. Recommender system MVP, an LLM evaluation harness, and an MLOps skeleton (rental prediction) for production discipline.
View projects → Get in touch
— 02 / Principles

Three rules.

Non-negotiable
01 / Proof

Demos, not decks

Every project ships with a live URL, screenshots from real runs, and a written evaluation. No "improved efficiency" claims without numbers attached.

02 / Rebuild

Reproducible by default

Notes list the secrets, endpoints, gotchas, and the exact order of operations. If you can't follow them to a working copy, they get rewritten.

03 / Status

Honest labels

Working MVP, In Progress, Planned — each one means what it says. Roadmap items are not pretended to be finished.

— 03 / Coverage

Areas covered across projects

Index of expertise
— AI / ML areas A
— Tools & stack B