Joseph Nawar APPLIED AI · MACHINE LEARNING
All selected work
PROJECT 01 / APPLIED AI + WORKFLOW ENGINEERING

OpsFlow AI

Purchase-order automation with structured extraction, deterministic validation, human approval, and resilient business-system integrations.

Explore the case study
Role
Software / AI engineering project
Stack
Python · FastAPI · PostgreSQL · n8n · React · Docker
Focus
Extraction · integrations · failure recovery
Illustrative pipeline from a purchase-order document to extraction, approval and Odoo, HubSpot, Gmail and Slack.

Original explanatory graphic, not a screenshot of the application.

LLM + rulesStructured extraction with deterministic validation
Human approvalReview before downstream synchronization
Audit trailsPersistence, retries, and recoverable state
01 / THE PROBLEM

Making document automation accountable.

Purchase-order workflows combine unstructured documents, business rules, approvals, and third-party systems. Extracting a value from a PDF is only one part of the task; the system also needs to validate that value, avoid duplicate actions, recover from failures, and make changes inspectable.

OpsFlow AI is my purchase-order automation project, built around document ingestion, structured LLM extraction, deterministic checks, human approval, and synchronization with operations systems.

02 / HOW IT WORKS

Keep uncertain steps separate from irreversible actions.

01 / INGESTReceive and parse purchase orders
02 / EXTRACTReturn structured fields via LLM
03 / CHECKRun deterministic rules and human approval
04 / SYNCIntegrate and audit downstream changes

The stack brings together Python, FastAPI, PostgreSQL, n8n, React and Docker. It integrates with Odoo, HubSpot, Gmail and Slack, using persistent state, idempotent synchronization, retries, recovery paths, and audit trails.

My emphasis was not simply getting the happy path to work. Integrations can timeout or replay requests; a reliable workflow needs to know whether an external action already happened, what remains pending, and when a person must approve it.

03 / EVIDENCE

Validation beyond a demo.

Work on the project included evaluation of extraction correctness, reliability, latency, and token cost, as well as automated testing. Its portfolio value is the engineering of the workflow and its failure-handling boundaries.

Evidence scope: The supplied CV describes implemented capabilities and evaluation areas, but does not report a public production deployment or numeric performance benchmarks. This case study does not claim either.
04 / DESIGN DECISIONS

Where judgment matters.

  • LLM vs. rules: use an LLM to handle variable document structure, while enforcing deterministic checks where possible.
  • Automation vs. oversight: include a human approval boundary before consequential downstream updates.
  • Simple sync vs. recoverable sync: prioritize idempotency, persistence, retries, and auditability when communicating with external platforms.

This is an engineering case study, not a claim that every possible business workflow is supported out of the box.

Next case study
Kawaneen ↗
View publications ↑
04 / LET'S CONNECT

Let’s build AI
that matters.

Hiring for applied AI or machine-learning engineering? I’m interested in remote-first teams, research collaborations, and carefully scoped engineering work. My inbox is open.

joe.hany2005@gmail.com

Elsewhere on the internet

Based in Cairo, Egypt.
Interested in remote-first opportunities.