AI systems engineer · available selectively

I build the systems I wish I had.

I came to software through operations, not theory. Now I build Python backends, AI workflows, and automation for messy work that needs to become reliable.

Yasser Saker
SHIP
THE
COMPLEX
31.95° N / 35.91° E
5+
years engineering
2M+
pages processed
7
years running a business
20+
AI bots in production
The useful context
NOT A STRAIGHT LINE

Before the APIs, there were customers, stock, and deadlines.

Yasser spent seven years running a construction materials business: procurement, suppliers, inventory, pricing, sales, and people. He learned software by solving the same kind of problems he was already responsible for.

01PracticalStart with the bottleneck.
02Systems-mindedMake the whole workflow work.
03End-to-endOwn it through production.
Selected systems

Less talking.
More proof.

Real products, real constraints, real delivery. Open a project for the architecture behind the demo.

01 / 04
WATCH / OPEN ↗
System 01

Social Responder AI

94%faster responses

Context-aware customer responses across messaging platforms — WhatsApp, Instagram, Facebook, Telegram, TikTok, X, Threads & web chat.

AI agentNLPAutomation
02 / 04
WATCH / OPEN ↗
System 02

OrderBot

MENU → PAYfull order flow

Conversational ordering from menu to payment — customers order inside WhatsApp or Telegram, you manage from a real-time dashboard.

Conversational AIPOSWhatsApp
03 / 04Restaurant ERP
OPEN ↗
System 03

Restaurant ERP

ORDERS → STOCKfull restaurant ops

Complete restaurant management — orders, menu, tables, inventory and staff — in one real-time system.

ERPPOSInventory
04 / 04Hermes · Harness
OPEN ↗
System 04

Hermes · Harness · Channels

AGENTconnected to Telegram & WhatsApp

Autonomous AI agent (Hermes) wired through the DeepSeek harness — connected to Telegram and WhatsApp for real conversations, automations and operations.

AI agentTelegramWhatsApp
How I work

Think like an operator.
Build like an engineer.

01

Map the real problem

Clarify the workflow, the failure modes, and the metric that matters.

↓
02

Design the spine

Choose the simplest architecture that can survive real traffic and real change.

↓
03

Ship the useful slice

Build a working path early, then harden it with tests, observability, and recovery.

↓
04

Make it boring to operate

Production is the product: deployment, queues, logs, and predictable maintenance.

↓
The short version

Not just an engineer.
An operator who codes.

Five years shipping software. Seven years running a real business. That combination matters: I understand the workflow before I automate it, and I stay close to the outcome after deployment.

Read the CV ↗
PythonFastAPI · Flask · Celery
AILLMs · RAG · LangChain
DataPostgreSQL · Redis · MinIO
InfraDocker · Linux · Nginx
Start a conversation

Bring me the messy part. I like it.

Available for selected backend, AI, and automation work.

yasser.saker95@gmail.com ↗
LinkedIn ↗

PROJECT DEMO