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