AI Automation & Python Developer focused on reliable LLM, agent and workflow systems.

I build practical automation around Python/FastAPI, REST APIs and webhooks, structured LLM outputs, workflow orchestration, browser/document automation, cloud deployment, validation, retries, idempotency and regression testing.

Have one reproducible Python bug? Request a fixed-scope $35 USD repair + regression-test pilot · Open the safe public issue form. No obligation or upfront fee to enquire; acceptance criteria, delivery and payment are agreed first. See current-head compiler-backed CI proof. These are independent open-source engineering examples, not client testimonials.

For founders & small-business operators: Free 4-question tool finder — find the right next step before paying for marketing software. No sign-up; recommendations are based on your business need, with affiliate disclosure when relevant.

Planning your first lead-capture funnel? Use the free interactive plan-fit calculator to compare actual contact, funnel, automation-rule and domain needs with the published Free and paid tiers. It takes no signup or payment; the independent guide labels its optional affiliate referral, and current checkout terms should be verified with the vendor.

Not sure which tool actually fits? See the five-scenario business software buyer-fit guide — independent comparisons, free-first buying checks, and clearly identified affiliate vs ordinary links.

Starting a newsletter or a lead-capture page? 2026 free email-vs-funnel decision guide — compares current vendor-published Free plans for systeme.io, MailerLite and Carrd, with practical tests before paying. The systeme.io link inside is disclosed as affiliated; MailerLite and Carrd links are ordinary vendor links. No product endorsement or conversion claim.

Need a client inquiry page set up rather than another software guide? I offer one $35 USD, written-only setup pilot: one inquiry page, its confirmation step and one basic follow-up check, after we agree on the brief and available features. No purchase or account subscription is included; delivery timing, access and PayPal payment are agreed before work begins. Request the written scope by email. Enquiries are free; this is not a checkout or an affiliate commission.

• AI workflow automation and integration
• Python + FastAPI
• REST APIs, webhooks and system handoffs
• LLM API integrations and structured output reliability
• Agent guardrails, bounded retries and stop conditions
• Pydantic v2 and JSON Schema validation
• Regression testing with pytest
• Docker / AWS Lambda / cloud deployment
• Workflow reliability, observability and failure recovery

A self-built public engineering proof for unreliable structured LLM output.

It demonstrates:

• safe JSON extraction without eval or fabricated repair
• strict Pydantic v2 validation
• explicit structured error states
• FastAPI endpoints
• Docker support
• AWS Lambda deployment
• 11/11 public regression tests passing

Live API docs:
https://7hjfrxuwyxgke2h7fqhaubykda0benzq.lambda-url.us-east-1.on.aws/docs

A runnable public proof for containing common agent failure modes before dispatch.

It focuses on:

• repeated identical tool-call detection
• bounded execution
• projected LLM-spend controls
• explicit stop conditions
• testable guard behavior

A self-contained example of idempotent single-batch contact imports, data validation, CSV formula defenses and audit-safe error reporting. 18/18 local Python regression tests passed (25 Oct 2026); not a live CRM integration or paid-client reference. Review the tested source and limitations.

A public portfolio of sanitized, real engineering and operations proofs.

Current case studies:

• Automation Reliability Proof — verified cloud-to-browser acceptance facts, serialization, persistent session continuity, fail-closed binding checks and regression gating.
• Connector-First Outreach Operations — research → enrichment → one-to-one outreach → booking CTA → reply/bounce monitoring, with clear boundaries between sent state and actual conversion.
• Recovery Engineering — evidence-led incident isolation that preserves known-good controls instead of rebuilding a healthy core.
• Deployment Runtime Verification — a real cloud deployment case showing why saved configuration, deployment status and actual serving runtime must be verified separately.
• LLM Structured Output Reliability — a focused case study on safe extraction, strict Pydantic contracts, explicit failure states and regression testing for model outputs.

Private implementation details, secrets, session identifiers and proprietary control-stack internals are intentionally not published.

• اختر أداة مشروعك في 4 أسئلة — اختبار عربي مجاني — أداة تفاعلية تساعد المستقلين وأصحاب المشاريع على تحديد الحاجة بين صفحة هبوط وجمع عملاء والبريد والأتمتة والصوت/الفيديو، مع توصيات حيادية قبل شراء أي اشتراك. لا تجمع بيانات اتصال؛ فقط رابط systeme.io الشخصي المؤكد يظهر عند الحاجة الملائمة، مع إفصاح.
• دليل عربي عملي: إنشاء صفحة مجانية لجمع طلبات العملاء — خمس خطوات ومثالان قابلان للتكييف لخدمات التصميم وصيانة الحواسيب، مع قائمة فحص مجانية بلا تسجيل. يحتوي الدليل على رابط أفيلييت مُفصح عنه وقد نحصل على عمولة عند شراء مؤهل.
• قائمة الفحص العربية المجانية لصفحة جمع العملاء — خمس خطوات لاختبار رسالة الصفحة والنموذج والتسليم على الهاتف، دون اشتراك أو شراء.

For small-business operators comparing funnel/email plans: See the official systeme.io plan comparison. This is a disclosed affiliate link: Askoure Systems may earn a commission on a qualifying purchase, at no extra cost to you. You can start with the free plan instead.

• شرح عربي: إنشاء صفحة مجانية لجذب العملاء باستخدام systeme.io — خطوات عملية بالعربية وقائمة تحقق مجانية لأصحاب المشاريع الصغيرة. يتضمن رابط إحالة مع إفصاح واضح.

• Open-source Landing Page Release Checker (Python) — a bounded, read-only 13-check command-line tool for page metadata, server rendering, sitemap discovery, affiliate disclosure, and external-link hygiene; includes executable offline fixtures.

• Landing Page Self-Audit — free interactive tool — a ten-point self-check of message clarity, mobile readiness, working forms, privacy and conversion measurement, with a personalized fix list. No login required.

• How to Create a Free Lead-Capture Page in systeme.io — independent step-by-step setup guide with a free downloadable copy-and-QA kit. The guide discloses its affiliate relationship.

• Créer une page de capture gratuite avec systeme.io (français) — guide pratique en français, exemples pour indépendants et entreprises de services, modèle et checklist gratuits. L'article précise les liens d'affiliation.

• 2026 Moroccan blue zellige & small-space decor guide — five renter-friendly wall-art ideas inspired by clearly dated Pinterest trend data, with accurate digital-only disclosures and the existing Etsy shop.

• Free 4×6 / 5×7 printable art printing guide — avoid accidental cropping, choose the matching JPG, understand DPI-versus-image-detail, and see the current ALFAROUQWorks Etsy shop. Digital downloads only; current shop availability should be checked on Etsy.

These are practical educational resources, not claims of measured conversion gains.

The public proofs above have packaged commercial versions for teams that want the runnable implementation, tests, documentation, and handoff material rather than rebuilding the reliability layer from scratch.

• LLM Reliability Rescue Kit — $19 — Python/FastAPI structured-output validation, strict Pydantic contracts, CLI tooling, regression corpus, tests, CI, and production guidance.
• Agent Runaway Guard Kit — $24 — local Python guardrails for repeated tool calls, cyclic loops, retry storms, token/API-cost boundaries, replay testing, and CI regression gates.

Both kits are one-time purchases and are intended for adaptation inside personal, internal, or client projects according to their included licenses.

1. Reproduce or measure one concrete problem.
2. Define the expected contract or outcome.
3. Isolate the smallest failing component.
4. Add the smallest reliable fix.
5. Make failures explicit instead of silently inventing data.
6. Lock behavior with focused regression tests.
7. Document the handoff and keep known-good controls intact.

One reproducible Python bug + focused fix + regression tests — $35 USD

Written, asynchronous delivery through a PR or patch. Scope, acceptance check, delivery date and payment route are agreed before work; PayPal preferred. Inspect two recent open-source contributions and their CI evidence. AI assistance is disclosed.

I am open to remote contract, paid pilot, retainer or full-time work involving:

• AI automation
• workflow orchestration
• Python/FastAPI
• APIs and webhooks
• LLM / agent integration
• reliability and regression testing
• implementation and technical handoff

For a small automation or reliability problem, I can start with one bounded workflow or failing step and define an acceptance test before expanding scope.

• Askoure Systems: https://askoure-systems.floot.app
• LLM Reliability Rescue Kit: https://mhamedasskoure.gumroad.com/l/llm-reliability-rescue-kit
• Agent Runaway Guard Kit: https://mhamedasskoure.gumroad.com/l/agent-runaway-guard-kit
• Automation Discovery: https://automation-discovery.floot.app
• ROI Calculator: https://automation-roi-calculator.floot.app
• Upwork: https://www.upwork.com/freelancers/~01f95546c501a4254e
• Fiverr: https://www.fiverr.com/mhamedaskoure
• LinkedIn: https://www.linkedin.com/in/mhamed-askoure-1b02b7438/

Current focus: AI Automation · Python · FastAPI · APIs/Webhooks · LLM Reliability · Agent Guardrails · pytest · Docker · AWS