Ayi NEDJIMI

Cybersecurity Expert & AI Engineer | OSCP. Active Directory, Azure/AWS/K8s, CUDA & LLM fine-tuning, C++23. 20+ yrs, 50+ audits, 100+ OSS. ORCID: 0009-0001-5056-3253. ayinedjimi-consultants.fr/articles

France Joined on May 21, 2026 https://ayinedjimi-consultants.fr

Pronouns

he/him

Work

Senior Cybersecurity Consultant & AI Engineer at Ayi NEDJIMI Consultants

Ayi NEDJIMI

May 22

Go Fiber vs Gin vs Echo: what I chose after building a 1,600-article platform

#go #webdev #programming #backend

5 min read

Want to connect with Ayi NEDJIMI?

Create an account to connect with Ayi NEDJIMI. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

Ayi NEDJIMI

May 22

I built a Telegram Twitter auto-poster — here's what OAuth 1.0a actually requires

#python #twitter #api #automation

5 min read

Ayi NEDJIMI

May 22

pfSense vs commercial firewalls: what I learned auditing 200+ networks

#security #networking #devops #selfhosted

4 min read

Ayi NEDJIMI

May 22

I indexed 500+ security terms into a searchable glossary — what I learned about knowledge bases

#webdev #database #ai #knowledge

5 min read

Ayi NEDJIMI

May 22

LLM output validation: 5 patterns that actually work in production

#python #ai #llm #programming

6 min read

Ayi NEDJIMI

May 22

A practical guide to prompt engineering for structured data extraction

#python #ai #llm #tutorial

6 min read

Ayi NEDJIMI

May 22

Fine-tuning vs RAG: a decision framework with examples

#ai #llm #machinelearning #python

6 min read

Ayi NEDJIMI

May 22

How to detect prompt injection attacks in user input

#security #ai #llm #python

7 min read

Ayi NEDJIMI

May 22

Building an AI agent with tool use in Python (from scratch, no framework)

#python #ai #agents #tutorial

6 min read

Ayi NEDJIMI

May 22

DNS security in 2026: 10 things developers get wrong

#security #dns #devops #webdev

5 min read

Ayi NEDJIMI

May 22

Kerberoasting for developers: why your Active Directory is probably misconfigured

#security #activedirectory #devops #career

4 min read

Ayi NEDJIMI

May 22

Building a real-time SEO scoring engine in Go: 8 rules that actually moved rankings

#go #seo #webdev #programming

5 min read

Ayi NEDJIMI

May 22

How to build a production RAG pipeline in Python (without a vector database)

#python #ai #llm #tutorial

5 min read

Ayi NEDJIMI

May 22

I automated PDF generation for 1,600 security guides — WeasyPrint lessons

#python #webdev #security #devops

4 min read

Ayi NEDJIMI

May 21

I built a search engine over 1,600+ cybersecurity articles — here's what I actually learned

#ai #meilisearch #webdev #security

4 min read

Ayi NEDJIMI

May 21

I used LLMs to rewrite meta descriptions for 1,600 articles — honest results

#ai #seo #webdev #llm

5 min read

DEV Community

Ayi NEDJIMI

Skills/Languages

Currently learning

Currently hacking on

Available for

Go Fiber vs Gin vs Echo: what I chose after building a 1,600-article platform

Want to connect with Ayi NEDJIMI?

I built a Telegram Twitter auto-poster — here's what OAuth 1.0a actually requires

pfSense vs commercial firewalls: what I learned auditing 200+ networks

I indexed 500+ security terms into a searchable glossary — what I learned about knowledge bases

LLM output validation: 5 patterns that actually work in production

A practical guide to prompt engineering for structured data extraction

Fine-tuning vs RAG: a decision framework with examples

How to detect prompt injection attacks in user input

Building an AI agent with tool use in Python (from scratch, no framework)

DNS security in 2026: 10 things developers get wrong

Kerberoasting for developers: why your Active Directory is probably misconfigured

Building a real-time SEO scoring engine in Go: 8 rules that actually moved rankings

How to build a production RAG pipeline in Python (without a vector database)

I automated PDF generation for 1,600 security guides — WeasyPrint lessons

I built a search engine over 1,600+ cybersecurity articles — here's what I actually learned

I used LLMs to rewrite meta descriptions for 1,600 articles — honest results