Mouad Jaouhari mouadja02

`~/ whoami` · the manifest, as an agent would read it

# mouad.agent.yaml — system manifest
identity:
  name:    Mouad Jaouhari          # alias: MJ
  role:    Data & AI Engineer
  company: ATOS
  motto:   "From signals to systems — now with a mind of their own."

mission: >
  Turn raw, messy data into autonomous systems that
  retrieve, reason, and act — not just dashboards that report.

operating_loop: [ plan, retrieve, act, observe, reflect ]

stack:
  data:    [ Snowflake, Spark, Airflow, Kafka, dbt, Iceberg ]
  agentic: [ LangGraph, LangChain, RAG, CAG ]
  memory:  [ Qdrant, Pinecone ]
  tooling: [ Claude Code, OpenAI Codex, Cursor ]
  cloud:   [ AWS, GCP, Cloudflare ]

currently_building: [ dacli, dag-doctor, CAG-Lab ]
status: "open to agentic-AI & data-platform collaborations"

How I think about agents — the reasoning loop

I don't build chatbots; I build systems that plan, pull their own context, take actions, and self-correct. Every agent I ship runs some version of this loop:

%%{init: {'theme':'dark','themeVariables':{'primaryColor':'#0d1117','primaryTextColor':'#e7e5e4','primaryBorderColor':'#00f0ff','lineColor':'#a855f7','fontFamily':'monospace','fontSize':'15px'}}}%%
flowchart LR
    U([ prompt ]) --> P[plan]
    P --> R["retrieve<br/>Qdrant · Pinecone"]
    R --> A["act<br/>tools · APIs · SQL"]
    A --> O[observe]
    O --> D{goal met?}
    D -->|no| Rf[reflect] --> P
    D -->|yes| Out([ answer / action ])