anupam@wagle:~$ whoami
machine learning engineer
● open to machine learning & AI security roles
anupam@wagle:~$ cat about.txt
I started in computer vision — teaching models to read handwritten math, sharpen blurry images, and spot pedestrians in LiDAR point clouds — and followed the field into NLP and large language models, where I now spend most of my time.
These days I'm doing an MS in Computer Science in the US and working as a graduate research assistant, digging into how ML systems work under the hood and how to build them responsibly. I like problems where the model has to meet the real world: messy inputs, tight constraints, actual users.
anupam@wagle:~$ ls skills/
anupam@wagle:~$ ls -la projects/
total 4
Write an equation on a canvas, get the solution — plus a graph. Handles single- and multi-variable equations end to end: recognition, solving, and visualization.
Upscales images to 2x and 4x without turning them to mush — preserving detail and sharpness instead of just stretching pixels.
Classifies malaria-infected blood cells and shows its work: class activation maps highlight exactly which regions drove the prediction.
3D object detection from LiDAR point clouds — finds pedestrians, cars, and cyclists, and measures their distance from the sensor vehicle.
anupam@wagle:~$ cat experience.log
graduate research assistant · University of South Dakota
Research in machine learning as part of the MS in Computer Science program.
machine learning engineer · icebrkr, Switzerland
Machine learning and nlp work on user behaviour tracking system
machine learning engineer · LogicTronix-AMD-Xilinx Partner
Machine learning and computer vision work on FPGA-accelerated systems.
anupam@wagle:~$ cat education.log
MS, Computer Science · University of South Dakota
B.E., Computer Engineering · Kathmandu Engineering College
+2 Science · National School of Sciences (NIST)
SLC · Triyog High School
anupam@wagle:~$ cat contact.txt
Email is the fastest way to reach me — I read everything. For code, find me on GitHub; for everything else, LinkedIn works too.
# this prompt actually works — type help and hit enter.