I am a Computer Science Engineering student at KIT – Kalaignarkarunanidhi Institute of Technology, Coimbatore, specialising in Artificial Intelligence, Computer Vision, and Machine Learning. I design and deploy real-world intelligent systems — from microplastic detection with YOLOv8 to precision agriculture platforms integrating aerial drones and IoT sensing.
My engineering philosophy centres on building systems that are accurate, scalable, and purposeful. I have validated this through national-level hackathon victories — winning ₹25,000 prizes twice and placing 4th among 435+ teams at the TNWISE State Hackathon. I bring a strong foundation in deep learning, LLM fine-tuning, and full-stack prototyping, complemented by hands-on experience with PyTorch, OpenCV, Raspberry Pi, and cloud-based analytics dashboards.
Open To:
- AI / ML Engineering Internships
- Research Collaborations (Computer Vision, Environmental AI)
- Open Source Contributions
- Hackathon Teams
- Technical Mentorship & Communities
Languages
Frameworks & Libraries
Backend & Databases
Cloud, DevOps & Tooling
| Domain | Proficiency | Details |
|---|---|---|
| Object Detection | ████████████ Expert | YOLOv8, real-time microplastic & contaminant detection at 83–96% accuracy |
| Convolutional Neural Networks | ████████████ Expert | Custom CNN architectures for image classification and feature extraction |
| LLM Fine-Tuning | ██████████░░ Advanced | Ollama-based domain-specific fine-tuning, prompt engineering, local inference |
| Precision Agriculture AI | ██████████░░ Advanced | NDVI-based crop stress analysis, IoT-driven soil monitoring via Raspberry Pi 5 |
| Supervised Learning | ████████████ Expert | Regression, classification, model evaluation — NPTEL Elite Certified |
| Unsupervised Learning | █████████░░░ Proficient | Clustering, dimensionality reduction, anomaly detection |
| Generative AI | █████████░░░ Proficient | Prompt engineering, GenAI applications — SAWIT.AI / GUVI Certified |
| Data Engineering | █████████░░░ Proficient | NumPy, Pandas, Matplotlib — preprocessing, augmentation, pipeline design |
🔬 Microplastic Detection System using Machine Learning
A portable, real-time device that detects, classifies, and counts microplastic contaminants in water samples using fluorescence imaging, Raspberry Pi processing, and YOLOv8-based machine learning — making invisible environmental threats visible and measurable at the point of source.
| Attribute | Details |
|---|---|
| Stack | Python · YOLOv8 · PyTorch · OpenCV · Raspberry Pi · Fluorescence Imaging |
| Performance | 83–86% accuracy (deployment model) · 93–96% accuracy (competition model) |
| Scale | Real-time inference on edge hardware |
| Impact | Winner — SCIMIT 26, MVIT Puducherry (₹25,000) · Runner-Up — 36-Hour National Hackathon, KIT Coimbatore (₹25,000) |
| Domain | Environmental AI · Computer Vision · Edge Deployment |
| Repository | github.com/Atchayasree03/MICROPLASTIC |
Engineered an end-to-end pipeline: fluorescence image acquisition → preprocessing → YOLOv8 object detection → contaminant classification → live dashboard reporting. The system is designed for field deployment with no cloud dependency, running entirely on Raspberry Pi hardware and providing actionable water quality metrics in real time.
🌾 SORA — Smart Omniphibious Precision Agriculture System
An AI-enabled precision agriculture platform combining an aerial drone and a detachable ground rover for comprehensive, real-time crop and soil monitoring — enabling data-driven agricultural decision-making at scale.
| Attribute | Details |
|---|---|
| Stack | Python · Raspberry Pi 5 · IoT Sensors · NDVI Analysis · Cloud Analytics · Live Dashboards |
| Performance | Multi-modal sensing: visual, spectral, and chemical soil/crop data streams |
| Scale | Field-deployable dual-mode system (aerial + ground) |
| Impact | Detects crop stress, pest infestation, soil moisture, pH, and NPK levels |
| Domain | Precision Agriculture · IoT · Computer Vision · Environmental Sensing |
| Security | Cloud-based encrypted telemetry and sensor data pipelines |
Integrated NDVI-based aerial crop analysis with ground-level IoT soil sensing across moisture, pH, and NPK parameters. The Raspberry Pi 5 acts as the on-device edge processor, streaming telemetry to cloud-based analytics dashboards for agronomic insight and early intervention alerts.
🧬 LLM Fine-Tuning using Ollama
A domain-specific LLM optimisation project focused on improving contextual response accuracy through fine-tuning, prompt engineering, and efficient local inference deployment using Ollama.
| Attribute | Details |
|---|---|
| Stack | Python · Ollama · Large Language Models · Prompt Engineering |
| Performance | Measurable uplift in domain-specific response accuracy post fine-tuning |
| Scale | Local inference — no external API dependency |
| Impact | Reduced hallucination rate; improved task-specific contextual understanding |
| Domain | NLP · Generative AI · Model Customisation · Local Inference |
Implemented end-to-end workflows for model customisation: dataset curation, fine-tuning configuration, prompt template design, and inference benchmarking. Focused on deploying performant LLMs in resource-constrained environments without reliance on cloud inference endpoints.
🐾 PawsConnectHub — Pet Community Platform
A web-based platform connecting pet owners, services, and communities — enabling adoption listings, lost-and-found tracking, and AI-powered pet care support.
| Attribute | Details |
|---|---|
| Stack | JavaScript · HTML · CSS · Web APIs |
| Domain | Full Stack Web Development · Community Platforms · AI Integration |
| Repository | github.com/Atchayasree03/Pawsconnecthub |
| Recognition | Event | Details |
|---|---|---|
| 🥇 Winner | SCIMIT 26 — MVIT College, Puducherry (2026) | AI-based microplastic analysis system · 93–96% accuracy · ₹25,000 prize |
| 🏅 4th Place | TNWISE Hackathon — KCT College, Coimbatore (2026) | Ranked 4th among 435+ teams statewide · innovative solution pitch |
| 🥈 Runner-Up | 36-Hour National Hackathon — KIT College, Coimbatore (2025) | ML model for plastic contaminant detection in water · ₹25,000 prize |
AI & Machine Learning
-Generative_AI_Certification-7C3AED?style=for-the-badge&logo=openai&logoColor=white){kind=icon}
| Provider | Certification | Focus Areas |
|---|---|---|
| NPTEL | Introduction to Machine Learning — Elite | Supervised Learning · Regression · Classification · Model Evaluation |
| Infosys Springboard | Artificial Intelligence Primer | AI · ML · Deep Learning · Computer Vision |
| SAWIT.AI / GUVI | Generative AI | GenAI Concepts · Prompt Engineering · AI-Driven Applications |
Learning:
- Advanced Computer Vision architectures (DETR, SAM, GroundingDINO)
- MLOps and model deployment pipelines (ONNX, TensorRT, FastAPI)
- Retrieval-Augmented Generation (RAG) systems
Building:
- Production-grade microplastic detection systems for water safety
- AI-powered precision agriculture platforms with IoT integration
- Domain-specific fine-tuned LLMs for environmental monitoring
Exploring:
- Multimodal AI (vision + language)
- Edge AI deployment on embedded hardware
- Federated learning for privacy-preserving AI
Open To:
- AI/ML Engineering Internships (Remote or Coimbatore)
- Research positions in Computer Vision or Environmental AI
- Open source collaboration on impactful AI projects
- Hackathon partnerships
"Intelligence is not about knowing everything — it is about building systems that learn what matters."