Hello, I’m

I’m a computer engineering student who wants to know what the machine is really doing. These days I keep coming back to Rust

In a hurry? Jump to the short version and résumé
scroll

How it started

When something works, I want to know why.

That habit is how I ended up on Arch, writing Rust, and poking at the Linux kernel through eBPF. Lately it has pulled me towards data: how it moves, how it breaks, and how a machine can learn something useful from it.

I’m in my third year of computer engineering. I write about what I build, and about what breaks along the way.

Where I studied

A diploma first, then a degree.

Both in computer engineering: a diploma first, then a B.E. Everything on this page is what I’ve picked up along the way, in class and out of it.

  1. Where it began

    Diploma in Computer Engineering

    Marathwada Mitramandal’s Polytechnic

    The fundamentals that everything since has been built on.

  2. Where I am now

    B.E. in Computer Engineering

    DY Patil College of Engineering and Innovation

    Still in progress, and pointed at data engineering, DataOps and machine learning.

    Third year

What I’ve been building

Three things I made to find out how they work.

sloth-daemon

Every Linux user has watched one runaway process eat the whole machine. The usual fix is to kill it. I wanted something gentler.

sloth-daemon watches CPU use through eBPF, asks a small local model running on Ollama what kind of process it is looking at, then slows it down with cgroup v2 instead of ending it. It’s written in Rust with aya.

Read the code
An illustration of the idea, not real output.

AcciVision-AI

A camera watches the road, and a CNN+LSTM model watches the camera. The CNN reads each frame, the LSTM keeps track of what happens across frames, and when the sequence looks like an accident the system raises an alert on a Flask monitoring dashboard.

It’s an end-to-end, real-time pipeline, from video in to alert out.

Read the code
A cartoon of what the model is looking for.

Warehouse Robot Path Planning

My third-year mini project asks a simple question: how should a robot cross a warehouse floor?

I implemented A*, Dijkstra and breadth-first search in Python and compared how each one explores the same floor. Watch them take turns on the right. The repository is on its way.

A*
A tiny replay of the idea, not the project’s own code.

Those are three of the bigger ones. The rest of my repositories are on GitHub

Where I’m headed

Data engineering and DataOps, with machine learning on top.

The plan is to build the plumbing data travels through, keep it reliable, and put models on top. Everything I read, build and write points that way.

Raw datamessy, and everywhere
Pipelinesmoved and cleaned, reliably
Modelstrained, then watched
Decisionssomething useful comes out

Tools I built my skills upon

Languages

  • Python
  • Rust

Data & ML

  • CNN+LSTM
  • Jupyter
  • Ollama
  • Flask

Systems & tooling

  • Linux
  • Arch
  • eBPF
  • aya
  • cgroup v2
  • Git
  • GitHub

Things I write down

I blog about what I’m learning.

  • New posts are on the way.

Say hello

If this sounds like your kind of problem, say hi.

Moeen

Computer engineering student · data engineering, DataOps and ML

The short version

Everything above, in thirty seconds.

Who I am

A third-year computer engineering student heading towards data engineering, DataOps and machine learning. Rust and ML hobbyist, regular blogger, Arch user.

Where I studied

B.E. in Computer Engineering at DY Patil College of Engineering and Innovation (current, third year). Diploma from Marathwada Mitramandal’s Polytechnic.

What I’ve built

sloth-daemon, a Rust and eBPF CPU governor that throttles runaway processes with cgroup v2. AcciVision-AI, real-time accident detection with a CNN+LSTM and a Flask dashboard. Warehouse Robot Path Planning, a comparison of A*, Dijkstra and BFS in Python.

What I work with

Python, Rust, eBPF, aya, Linux, cgroup v2, Flask, Jupyter, Ollama, Git and GitHub.

Find me

github.com/mshkh-dev