Open to Junior AI / ML & Backend Roles
Aditya Gaharwar
>Agentic AI Workflows (LangGraph)
Local time--:--:--IST · India
I build AI systems using Python, modern LLM tools, and agentic frameworks. Most of my recent work focuses on stateful agent search pipelines, production RAG workflows, and the backend APIs needed to run them reliably. I care about building systems that are grounded, accurate, and ready for production.
Work Experience
Where I've applied my skills so far
AI Engineer
July 2025 – July 2026Radials International Mining Services Private LimitedDehradun | Telangana
- Visual Truck Inspection: Built an image analysis pipeline using the Google Gemini API to inspect 14 key points on trucks from photos. Set up strict response formats (JSON schemas) so data could be saved cleanly into our database, cutting manual review time by 80–90%.
- Backend & Database: Wrote backend endpoints in Python using FastAPI. Created SQL queries and Supabase functions (RPCs) to take sensor data and calculate maintenance alerts.
- Attendance Verification: Set up a face-recognition script that checks photo EXIF metadata (timestamp and location) to catch attendance fraud, saving about 60% of manual auditing time.
- Speed Improvements: Replaced sequential image processing with parallel batches in Python, reducing data processing time from hours to minutes.
PythonFastAPIGoogle Gemini APISupabasePostgreSQLFace RecognitionEXIF Validation
Research Intern
May 2024 – July 2024IIIT DharwadRemote
- Dataset Creation: Cleaned and labeled 1,000 unpaved road images using CVAT for an image segmentation model.
- Handling Uneven Data: The background took up most of the pixels compared to the actual roads. Used data augmentation and a custom PyTorch loss function to help the model learn road boundaries better.
- Model Training: Trained a DenseNet-121 model in PyTorch, reaching a 0.81 IoU score and 98% validation accuracy on our test set.
PythonPyTorchDenseNet-121CVATComputer Vision
Skills
Tools and technologies I work with regularly
AI & LLM Tools
- >LangGraph
- >LangChain
- >RAG Systems
- >Google Gemini API
- >Prompt Engineering
Machine Learning & Vision
- >Python
- >PyTorch
- >Scikit-learn
- >OpenCV
- >Face Recognition
- >Data Augmentation
Databases & Vector Search
- >Qdrant
- >PostgreSQL
- >Supabase
- >Redis
Backend & DevOps
- >FastAPI
- >Flask
- >REST APIs
- >Docker basics
- >Git
- >Linux
Featured Projects
Selected public work with interactive case studies