class HardikSankhla:
name = "Hardik Sankhla"
location = "Jodhpur, Rajasthan 🇮🇳"
education = "B.Tech Data Science @ JIET Institute of Design & Technology"
role = "AI Intern @ Gyan Netra"
currently = ["Learning Generative AI", "Exploring Agentic AI", "Building AI-powered tools"]
interests = ["Large Language Models", "Computer Vision", "Cloud Computing (AWS)", "Deep Learning"]
ask_me_about = ["Python", "ML/DL", "AWS", "Data Science", "Frontend Dev"]
fun_fact = "I love chess and solving logic puzzles — same skillset as debugging 🤣"| 🚀 | Activity |
|---|---|
| 💼 | AI Intern at Gyan Netra — building Generative AI pipelines |
| 📚 | Deep-diving into Agentic AI and multi-agent systems |
| 🧠 | Exploring LLMs, RAG, and LangChain in real applications |
| 🎓 | Final year B.Tech — Data Science @ JIET Jodhpur |
| 🌱 | Advanced Deep Learning — CNNs, Transformers, Diffusion models |
|
CNN · TensorFlow · Keras · Kaggle Built a Convolutional Neural Network that identifies dog breeds from images. Used Kaggle API for dataset acquisition, custom CNN architecture, and achieved strong test accuracy. |
LangChain · LLMs · Python Experimenting with autonomous AI agents — building multi-step reasoning pipelines that can plan, act, and self-correct using LLM backends. |
💡 More projects on my GitHub repositories →
- Working on Generative AI pipelines and LLM-based applications
- Exploring Agentic AI frameworks for autonomous task completion
- Managed AWS infrastructure: EC2, EBS, S3, EFS, snapshots
- Worked with AWS AI services: Rekognition, Comprehend, Transcribe, Translate, Forecast
- Learned monitoring, scaling, security, and automation
- Built responsive UIs with HTML, CSS, Bootstrap, JavaScript
- Focused on user engagement optimization
🏛️ B.Tech in Data Science (CSE) — JIET Institute of Design and Technology, Jodhpur (2022–2025)
This repository powers a multi-surface ecosystem:
- GitHub profile source in
profile/README.md - Portfolio source in
apps/web - Docs source in
apps/docs
For professional update workflow and AI agent operating rules, read: