Python Automation, Web Scraping & Backend Engineer
Python engineer building production-grade automation systems, backend APIs, and data acquisition pipelines — with a peak delivery of a million records in a day from direct protocol access, no browser, no DOM.
01 / About
I build systems by understanding the layers they depend on—from TLS fingerprints and browser protocols to API boundaries, persistence, and CI. The result is automation that keeps working when the easy assumptions stop being true.
Self-taught through the problems themselves: protected clients that had to be understood before they could be automated, and services that had to survive their own success. The payoff is measurable — on one engagement a peak day returned a million records from about 10,000 requests, because the work went to the protocol instead of the page.
I work well inside a team — collaborating with data scientists and ML engineers on delivery, taking direction, and applying feedback. That's reflected in my verified references.
4
Open-source tools
Public repositories with CI
2
Research articles
Published technical deep dives
1
Responsible disclosure
CWE-835 reported to Samsung
3
Verified references
Independently collected via ZincWork
600+
Targets automated
Sites and APIs across client work
2
Languages
Arabic native · English professional
02 / Capabilities
Tools are selected for control, observability, and a clean path from investigation to dependable production software.
Typed, observable services with clean data boundaries.
Browser and request-level systems built for resilience.
Protocol-aware data acquisition beyond the happy path.
Recovering behavior from protected web and mobile clients.
Repeatable delivery, tests, and operational foundations.
03 / Experience
Contract engagements rather than salaried roles: investigating difficult protocols, designing maintainable implementations, and carrying them into production. Embedded in small client teams alongside data scientists and ML engineers, delivering against their requirements and review.
2022 — Present
Freelance · Contract
Cairo, Egypt · Remote
Automation · Data acquisition · Reverse engineering · Backend APIs
Aggregate figures across client engagements from 2022 onward. Volume comes from reading the protocol directly rather than driving a browser.
1Mrecords
Peak day
Delivered from ~10K direct HTTP requests — roughly 100 records per call, no browser or DOM rendering.
600+targets
Sites and APIs
Distinct scrapers built against production targets across varied protections and response formats.
60GB35M rows
Largest delivery
Single dataset acquired, validated, and handed off in a structured format.
30providers
Challenge types
Verification and bot-mitigation flows handled on client-authorized targets, within contracted scope.
6-24hrecovery
Defense changes
Typical turnaround to diagnose and ship a fix when a target changed its protections.
95%+accuracy
capture2tls output
Share of captures converted to working request code without manual correction.
Figures span client engagements between 2022 and 2026. Client names, target systems, and delivered datasets are covered by confidentiality agreements; specifics are available on request under NDA, and the referees below worked directly on this delivery.
04 / Selected work
Each project isolates a hard systems problem, makes the architecture legible, and leaves the implementation open for inspection.
/01
A HAR-based HTTP provenance analyzer that traces the cookies, tokens, headers, and request dependencies behind a target request.
/02
A zero-dependency converter from raw proxy captures to fingerprint-aware Python request code.
/03
A pattern-based JavaScript deobfuscator that survives regenerated identifiers, alphabets, and rotation offsets.
/04
A FastAPI service for authenticated API integrations and background synchronization workflows.
05 / Research
Technical research is part of the engineering process: isolate assumptions, instrument the target, build the smallest useful model, and document what survives scrutiny.
Protected web systems
Mapped the layered interaction between generated JavaScript, device fingerprinting, behavioral telemetry, and server-side challenge verification.
Dynamic analysis · AST tooling · Headless reproduction
Mobile imaging security
Traced a non-terminating execution path in the UltraHDR stack and validated the denial-of-service condition through coverage-guided fuzzing.
arm64 reversing · Frida · AFL++
06 / Articles
Long-form field notes on the constraints, false starts, tooling, and mechanisms behind the result.
07 / References
Independently collected and identity-verified through ZincWork. Each referee worked directly with me on delivery.
“Based on my experience working with him, I would gladly collaborate with him again.”
He was responsible for collecting and organizing data, performing analysis, solving technical challenges, and ensuring tasks were completed on schedule.
“I would feel confident having him on the team again because of his professionalism, reliability, and collaborative approach.”
His role involved managing technical tasks, analyzing project requirements, contributing to problem-solving, and making sure his assigned work was completed accurately and on time.
Full reference pack, independently verified.
View on ZincWork08 / Credentials
Public code, CI history, and published research carry the engineering evidence. What follows is the formal record behind it.
Cairo University
Two things carry over. The first is practical: I can read a terms of service, a data processing agreement, or a scraping engagement scope and say where the line is before any code is written — which matters in this field. The second is the habit itself — close reading under adversarial conditions, where rules, exceptions, and edge cases decide the outcome. A protocol specification reads the same way.
Academic reference: Ayman Al-SaadanyAssistant Professor of Public Law, Cairo University
09 / Contact
Open to backend and automation roles, and to technically demanding contract work across data acquisition, reverse engineering, and Python backend systems.