Hi, I'm Prabesh Sharma.
Computer Engineer specializing in AI & ML solutions , building Models, Agentic systems, and Computer-vision pipelines that scale from research to real users.
AI Engineer
Prabesh Sharma
shippingEngineering AI that ships.
I'm a Computer Engineer who builds AI across computer vision, NLP, and LLMs. I train models, deploy them on the cloud, and run them in production, monitoring how they perform, debugging issues, and fixing them to keep things reliable. I've also Work on AI systems that automate the complete software development lifecycle from generating production codebases to deployment.
$ cat impact.json
Projects Shipped
Years of Engineering
Models in Production
Blog Posts Published
$ echo "philosophy"
// Models that ship beat models that win benchmarks. I optimize for systems that earn trust in production.
Production-grade AI
I treat models as products — measured, monitored, and built to outlive the demo.
End-to-end ownership
From data and architecture to APIs and deploys — I ship the whole stack.
Agentic thinking
Multi-agent design, tool use, and retrieval pipelines that actually work under load.
Research × engineering
Comfortable reading papers and translating them into systems users touch.
Experience that compounds.
From research labs to AI-first startups — every role taught me something I now use to build production systems.
AI Engineer
Inspiring Lab· Full-time
Building production AI across conversational systems, document intelligence, computer vision, and data pipelines from prototype to monitored deployment.
- Developed an AI-powered sales chatbot with intent recognition, semantic product search, multi-format document ingestion, shopping…
Software Engineer — AI
Skillrank· Full-time
Built an AI-powered SDLC automation platform integrating multimodal LLM workflows and agentic systems to scaffold and deploy full-stack projects.
- Developed an AI-powered SDLC automation platform integrating multimodal LLM workflows and agentic systems.
AI/ML Engineer Intern
Scala Technologies· Internship
Built production ML models spanning medical imaging, route optimization, conversational AI, and demand forecasting.
- Developed 2D and 3D models for brain tumor segmentation.
Projects with real impact.
A mix of research, production systems, and weekend builds. Tap any card to read the full case.
More projects
Skills built through shipping.
Not a list of buzzwords — these are tools I reach for daily, with confidence scaled to actual usage.
AI & Machine Learning
Where I spend most of my time.
Currently exploring
Education & curiosity.
Formal training that planted the roots — combined with a relentless habit of building beyond the syllabus.
2021 → 2025
Bachelor's Degree in Computer Engineering
Tribhuvan University, Institute of Engineering (TU, IOE)
Focused on software development, artificial intelligence, and machine learning.
Coursework
Self-directed
Beyond the classroom
- Published research on real-time anomaly detection using fine-tuned Vision Transformers.
- Active writer on Medium — explaining transformers, ML fundamentals, and SQL for data science.
- Mentoring AI interns building reinforcement-learning agents that learn from human feedback.
- Open-source contributor across LLM tooling, RAG pipelines, and computer-vision projects.
Notes from the lab.
Long-form explanations of the ideas I work with — written to make complex topics actually click.
Image Captioning Using Transformer
After numerous attempts with RNNs, GRUs, and LSTMs to generate captions for images, here's how transformers changed the game.
Machine Learning Algorithms
Machine Learning is a subset of artificial intelligence mainly concerned with the development of algorithms that learn from data.
Database and SQL for Data Science with Python
SQL — the language used for relational databases — meets Python: querying data, joining tables, and feeding pipelines.
Let's build something extraordinary.
An agent platform, a model to train and fine-tune, a custom LLM pipeline, a Python-powered backend, or a computer-vision system that sees what others miss — if it's ambitious and a little crazy, I'm already in.
Liked what you saw? Fuel the next build.
🥟 Buy me a momo