Muhammad Ibrahim — Portfolio

Senior Full-Stack Engineer · Production Go microservices · AI-accelerated development workflows

Highlights

Area	Evidence
Full-stack / platform	3-service arbitrage platform (TA, DMS, TES) — 1,400+ commits, gRPC, Redis, 5 exchanges
AI-augmented delivery	Cursor-driven dev workflow, offline E2E standards suite, PM→Dev→QA agent loop (portfolio)
Scale	100K+ DAU platform work (Aylab), cost optimization, ML fraud detection

Repositories (private — available on request)

Service	GitHub	Role
TA — Trading Agent	trading-agent	Real-time market ingestion & opportunity detection
DMS — Decision Making Service	decision-making-service	6-gate decisions & execution orchestration
TES — Trade Execution System	trade-execution-system	Multi-exchange order placement & fill tracking

Agent orchestration layer (this repo)

Learning & portfolio project: multi-role agent workflow on top of the trading codebase.

• Knowledge base — system docs for AI context (RAG-ready)
• E2E standards — run_offline.sh — 8 automated checks (no live exchanges)
• Agent prompts — PM, Dev, Reviewer, QA, E2E roles
• Cursor workflow — CURSOR_STEP3_PROMPT.md

You → PM (Cursor) → Dev (Cursor) → QA/E2E (run_offline.sh) → PASS

Start here: agent-orchestration/NEXT_STEPS.md

Architecture

TA (ingestion) ──gRPC──► DMS (decide) ──HTTP──► TES (execute)
         │                      │                      │
         └──────────── Redis tde:* lifecycle ──────────┘

Local paths (development)

~/desktop/trading-agent/              # TA (arbitrage-realtime inside)
~/Desktop/decision-making-service/    # DMS
~/Desktop/trade-execution-system/     # TES
~/Desktop/my-portfolio/               # Portfolio + agent orchestration