My entire academic and technical life — connected, searchable, and alive in one knowledge graph.
I finished my final year of university with more knowledge than I could track.
Real AI systems built and deployed. Semesters of theory studied and applied. Certificates earned and forgotten in PDF folders. Notes scattered across dozens of files.
None of it was connected. None of it was queryable. None of it was accumulating into anything useful.
Ask me "which of your projects prove your ML skills?" I would have to manually search five different folders.
Ask me "what did you learn in 3rd semester that you later used in production?" I had no way to answer that.
Ask me "which technology do you use most consistently across all your work?" I genuinely did not know.
That is the problem this wiki solves.
Every project, course, certificate, and concept connects back to a single master index node. Nothing is isolated. Everything is reachable.
The purple index node sits at the center.
Every line you see is a real relationship —
a project that uses a technology, a course that
explains a concept, a certificate that maps to
real code written and shipped.
The personal-portfolio node connects to every
project and credential worth showing publicly.
The wiki knows exactly what belongs here and why.
Every node in this graph is real work. Not placeholders. Not templates. Built, studied, earned, and documented.
Projects link to the technologies they use. Technologies link to the courses that taught them. Courses link to the semesters they belong to. Certificates link to the projects that prove them.
The muzzamil-khalid entity node sits beside index
as a personal identity anchor inside the wiki.
Every project I authored, every skill I demonstrated,
every credential I earned — connects back here.
This is not just a portfolio. It is a queryable record of everything I am as an engineer and learner.
Real AI systems. Not tutorials. Not clones. Original systems built to solve real problems.
- Pymentor AI — AI tutor platform with RAG
- AVA WhatsApp Agent — multi-agent WhatsApp assistant
- PropVideo AI — real estate video generation
- Legal Agentic Workflow — AI legal document system
- Multi-AI Finance Agent — autonomous finance analysis
- Automatic Job Apply System — agentic job applications
- AI Blogging Automation — end-to-end content pipeline
- AI Anime Recommendation — personalized rec system
- AI Travel Planner — intelligent trip planning
- Programmatic Video Generation — automated video creation
- Social Media Automation — cross-platform automation
- Automated Content Generation — multi-channel content AI
- Web Scraping Intelligence — smart data extraction
- Computer Vision Classifier — object detection system
- Crypto Prediction System — ML price forecasting
- Agentic Automation Workflows — multi-step AI pipelines
- Google Generative AI Projects — Gemini integrations
- Recommendation Systems — collaborative filtering engine
- ATS Resume Builder — AI-powered resume optimization
-
- more added continuously
Every semester. Every subject. Every concept. Not just stored — connected to the real projects that applied that theory in production code.
- 3rd, 4th, 5th, 6th, 7th, 8th semester courses
- Artificial Intelligence
- Machine Learning
- Data Structures and Algorithms
- Database Systems
- Computer Networks
- Operating Systems
- Software Engineering
- Web Development
- Probability and Statistics
- Computer Vision
- Discrete Mathematics
- Numerical Methods
- Final Year Project research and documentation
Every certificate I earned is linked directly to the projects that prove I applied that skill in real working systems.
Not just credentials sitting in a folder. Evidence of actual competency.
Papers read. Tutorials completed. Ideas explored. Concepts learned outside the classroom. All filed. All connected. All searchable.
A traditional portfolio answers one question: "what did you build?"
This wiki answers questions nobody else can answer about themselves:
-
"Which of my projects use RAG — and what university theory explains the architecture decisions I made?"
-
"Which certificate skill did I prove across the most real projects in production?"
-
"What patterns appear consistently across all my agentic systems?"
-
"What did I study in 2nd year that I later shipped in a production system?"
-
"Which technology do I have the deepest evidence for across my entire body of work?"
The difference is not storage. It is connection.
Every project added makes every other project richer. Every course linked to a project adds depth. Every certificate connected to real code becomes proof. Every note filed becomes permanently searchable.
Nothing resets. Nothing gets lost. Knowledge compounds.
Everything I build, learn, and earn
↓
Raw sources stay private and local
↓
Gemini CLI reads and processes
↓
Wiki pages written in markdown
↓
Obsidian renders the knowledge graph
↓
Any LLM tool queries it via MCP
Every time I build a new project — wiki updates. Every time I complete a course — wiki updates. Every time I earn a certificate — wiki updates. The graph grows. The connections deepen.
This wiki is built using the LLM Wiki pattern published by Andrej Karpathy on April 4, 2026.
Karpathy identified exactly the problem I had:
"Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation."
His solution — instead of retrieving from raw documents every time, an LLM builds and maintains a persistent wiki that compounds over time:
"The wiki is a persistent, compounding artifact. The cross-references are already there. The contradictions have already been flagged. The synthesis already reflects everything you've read."
And the reason it works:
"The tedious part of maintaining a knowledge base is not the reading or the thinking — it's the bookkeeping. LLMs don't get bored, don't forget to update a cross-reference, and can touch 15 files in one pass."
His pattern. My four years of work. The result is this wiki.
wiki/
├── index.md ← master catalog, every node listed
├── log.md ← full history of every update
├── sources/ ← one deep page per project/course/record
├── concepts/ ← technology and theory concept pages
└── entities/ ← frameworks, tools, people, institutions
Every source page includes:
- Full summary of what it is
- Key concepts and architecture
- Cross-references to related projects
- Links to concepts and entities it uses
- Evidence of skills applied
Browse on GitHub
Start at wiki/index.md for the full catalog.
Every page readable as plain markdown.
Clone and open in Obsidian
git clone https://github.com/Maimuzamilhu/personal-wikiOpen in Obsidian to see the full interactive knowledge graph — every connection visible.
- Gemini CLI — wiki maintenance agent
- Obsidian — knowledge graph visualization
- LLM Wiki pattern by Andrej Karpathy
Raw source files — actual code, certificates, personal documents — are excluded via .gitignore and never pushed to this repo.
This wiki contains only synthesized knowledge: summaries, architecture docs, concept pages, and cross-references.
This wiki grows every time I build or learn. It is never finished. That is the point.