AI Engineer building LLM-powered software systems with a focus on agentic workflows, backend APIs, and reliable infrastructure.
My work spans Python, FastAPI, data systems, model integration, workflow automation, and observability. I care about making systems not only functional, but testable, maintainable, and practical to deploy.
Focus areas:
- LLM applications, retrieval, and evaluation
- Agentic systems and tool integrations
- Backend engineering and asynchronous processing
- AI infrastructure, observability, and deployment
Selected work: Portfolio · GitHub repositories · LinkedIn
B.Tech in Computer Science & Engineering (Data Science)
JIET Institute of Design & Technology, Jodhpur
2021 – 2025 • Degree awarded
Sep 2025 – Feb 2026 | Jodhpur, India
- Paid internship (upgraded from unpaid internship)
- Building LLM applications, agentic workflows, and backend infrastructure
- Stack: FastAPI, PostgreSQL, Redis, RabbitMQ, Celery, OpenTelemetry, Docker, AWS
Nov 2024 – Aug 2025 | Jodhpur, India
- Phase 1 (Nov 2024 – Apr 2025, unpaid): Self-driven FastAPI backend development, CRUD APIs, database design
- Phase 2 (May 2025 – Aug 2025, paid): Production debugging, AI-assisted development, observability (LGTM stack), Redis/RabbitMQ/Celery integration
- Stack: FastAPI, PostgreSQL, Redis, RabbitMQ, Celery, OpenTelemetry, Docker, AWS
Apr – Sep 2025 | Unpaid
- Agentic AI patterns: AutoGen, LangGraph, CrewAI, AutoGPT
- Multi-agent orchestration, RAG pipelines, evaluation frameworks
Jul – Sep 2023 | Paid Training
- Infrastructure: EC2, EBS, S3, EFS, snapshots, IAM, VPC
- AI Services: Rekognition, Comprehend, Transcribe, Translate, Forecast
- Monitoring, scaling, security, automation
May – Jun 2023 | Paid Training
- Responsive UIs: HTML, CSS, Bootstrap, JavaScript
- User engagement optimization
Python FastAPI SQL JavaScript HTML/CSS Git Linux Docker
PyTorch TensorFlow scikit-learn Pandas NumPy LangChain Ollama ChromaDB Streamlit OpenTelemetry
AWS (EC2, S3, EBS, EFS, Lambda, Rekognition, Comprehend) PostgreSQL Redis RabbitMQ Celery Nginx
Kubernetes Kafka Grafana Loki Prometheus MLflow DVC Airflow Prefect
| Project | Description | Stack | Repo |
|---|---|---|---|
| relay-connect | Service communication layer | Python, FastAPI | github.com/Hardik-Sankhla/relay-connect |
| OpportunityOS | Opportunity tracking system | FastAPI, PostgreSQL | github.com/Hardik-Sankhla/OpportunityOS |
| AiOS | AI operating system concepts | Python, Streamlit | github.com/Hardik-Sankhla/AiOS |
| SCHOOL-OF-AI | Coursework & experiments | Python, LangGraph, AutoGen | github.com/Hardik-Sankhla/SCHOOL-OF-AI |
| CNN-Plastic-Waste-Classification | Deep Learning (CNN) | Python, TensorFlow, Keras, Jupyter | github.com/Hardik-Sankhla/CNN-Plastic-Waste-Classification |
| Mastering-Agentic-Design-Patterns | Agentic AI Coursework | Python, AutoGen, LangGraph, CrewAI | github.com/Hardik-Sankhla/Mastering-Agentic-Design-Patterns-with-Hands-on-Projects |
| PREDICT-EMPLOYEE-RETENTION | ML Classification | Python, scikit-learn, Jupyter | github.com/Hardik-Sankhla/PREDICT-EMPLOYEE-RETENTION |
| HOUSE-PRICE-PREDICTION | Regression | Python, scikit-learn, Jupyter | github.com/Hardik-Sankhla/HOUSE-PRICE-PREDICTION |
| Message-Using-Python | Twilio SMS Automation | Python, Twilio API | github.com/Hardik-Sankhla/Message-Using-Python |
The following projects are documented in this repository's canonical content (/content/projects/) and rendered on the portfolio website and documentation site:
| Project | Description | Stack |
|---|---|---|
| AI-Powered Chatbot | Context-aware support assistant with retrieval, memory, and safety guardrails | Python, LangChain, OpenAI, FastAPI |
| Computer Vision System | Real-time quality inspection pipeline with YOLOv8 and OpenCV | Python, YOLOv8, OpenCV, TensorFlow |
| LLM Evaluation & Observability Pipeline | Evaluation and telemetry framework for LLM releases | Python, FastAPI, OpenTelemetry, PostgreSQL |
| ML Recommendation API Platform | Personalized recommendation service with low-latency APIs | Python, Scikit-learn, FastAPI, Redis |
| Predictive Maintenance MLOps System | Predictive maintenance pipeline with automated retraining | Python, XGBoost, Airflow, Docker |
- Email: datascientist.hardiksankhla@gmail.com
- LinkedIn: linkedin.com/in/hardik-sankhla
- GitHub: github.com/Hardik-Sankhla
- PeerList: peerlist.io/hardiksankhla
- Location: Jodhpur, Rajasthan, India (IST)
"I build systems that think, models that evolve, and algorithms that predict."