Diya Chanda Logo
Diya Chanda
Open for AI Research & Engineering
B.Tech CSE (AI & ML)

Diya Chanda

>

Undergraduate researcher and engineer specializing in explainable deep learning, multi-modal sensor architectures, and precision agricultural AI. Published IEEE & Springer author dedicated to building high-accuracy, interpretable neural systems.

Explore Selected WorksDownload ResumeLet's Connect
4

Published Papers

7

Honors & Certs

4

Engineered Systems

Background & Philosophy

About Me

I am a B.Tech Computer Science & Engineering undergraduate specializing in Artificial Intelligence and Machine Learning at The Neotia University, maintaining a 9.48 CGPA. My work bridges empirical deep learning research with production-grade engineering, focusing on explainable computer vision, multi-modal neural architectures, and intelligent systems.

Diya Chanda — AI Researcher & Machine Learning Engineer

Diya Chanda

Artificial Intelligence & Machine Learning

Batch 2023 – 2027

Academic Profile

InstitutionThe Neotia University
Cumulative GPA9.48 / 10.0
Top Decile
Research FocusExplainable Deep Learning & Computer Vision
LocationWest Bengal, India

Interests & Explorations

Deep Learning ResearchGrounded AI & RAGComputer Vision & Grad-CAMAlgorithmic Problem SolvingPrecision Agriculture AIFull-Stack Web SystemsBadmintonEmerging Tech Exploration
Technical Pillars

Core Engineering & Research Focus

Generative Systems

Grounded AI & RAG Architectures

Architecting reliable, context-aware retrieval-augmented generation pipelines that ground model responses in verified domain knowledge while mitigating hallucinations.

Explainable AI (XAI)

Deep Learning & Vision Research

Designing multi-headed CNNs, Vision Transformers (ViTs), and Grad-CAM interpretability layers that provide transparent, visual decision rationales for mission-critical diagnostics.

Model Optimization

Precision Edge-AI Systems

Engineering lightweight neural networks with over 30% parameter reduction, optimized for low-latency inference on agricultural and healthcare edge sensors.

Production Systems

Full-Stack ML Engineering

Transforming algorithmic research into production-grade, interactive web applications using Next.js, FastAPI, Streamlit, and cloud microservices.

Trajectory

Academic & Research Milestones

2023 – Present

B.Tech CSE (AI & ML) — The Neotia University

Commenced undergraduate studies focusing on core computer science foundations, mathematics for machine learning, data structures, and neural network architectures (Current CGPA: 9.48).

2024 – Present

Empirical AI Research & Conference Publications

IEEE & Springer Co-Author

Authored and published 4 peer-reviewed papers across IEEE ICRITO, IEEE COMPUTINGCON, and Springer LNNS, developing multi-task crop and fruit grading models.

2025 – Present

Production-Grade AI Platforms & Hackathons

Smart India Hackathon & Industry Internships

Selected for SIH 2024 (1st round), completed engineering internships at Xeta Labs and DataSpace Academy, deploying scalable web applications and AI APIs.

Engineering Mission
"To engineer interpretable, high-accuracy intelligent systems that seamlessly transition from academic research into robust real-world production environments."
Guiding Principles
Empirical Research RigorGrounded AI & ReliabilityInterpretability & XAIClean Software ArchitectureContinuous Intellectual Curiosity

Architecture & Engineering Tooling

Technical Stack & Capabilities

A curated directory of frameworks, libraries, runtime environments, and infrastructure tools utilized across deep learning research and full-stack systems engineering.

AI/ML & Deep Learning

Research Core
5 tools

Multi-headed CNNs, Vision Transformers, Grad-CAM interpretability, and scientific computing.

PyTorchTensorFlowKerasscikit-learnOpenCV

Generative AI & NLP

Grounded Systems
4 tools

Context-grounded RAG pipelines, semantic embeddings, vector indexing, and LLM orchestration.

LangChainRAGLLMsVector Search

Programming Languages

Algorithmic Foundation
4 tools

Object-oriented architectures, low-level data structures, declarative queries, and backend scripting.

PythonJavaSQLC
Core Engineering Synthesis

Deep Learning Research

PyTorch & TensorFlow workflows optimizing multi-task CNNs and ViTs with Grad-CAM visual interpretability.

Grounded Retrieval (RAG)

LangChain and vector similarity pipelines ensuring verified knowledge retrieval and hallucination mitigation.

Full-Stack Systems

FastAPI, React.js, PostgreSQL, and Dockerized microservices architected for low-latency asynchronous throughput.

Production Platforms & Deep Learning Systems

Featured Systems & Projects

Enterprise compliance platforms, climate-smart agronomy engines, generative AI assistants, and explainable computer vision architectures.

CampusSphere — Enterprise Accreditation & Institutional Credit Banking System
Live Production Platform
Enterprise Full-Stack Platform
Index 01Case Study

CampusSphere

Enterprise Accreditation & Institutional Credit Banking System

Multi-tier enterprise compliance and credit banking platform for higher education institutions. Streamlines student co-curricular verification via a two-stage pipeline (Faculty Advisor Stage 1 & Admin Stage 2), dynamic Credit Policy Engine mapped to NAAC Criteria 1–7, cryptographic public credential verification (/verify/[id]), and 1-click NAAC/NIRF CSV compliance reports.

Next.js 15React 19Node.jsExpressPostgreSQLSupabaseUpstash RedisCloudinaryOpenTelemetryTailwind CSS
Launch Live AppGitHub Repo
JalDrishti (जलदृष्टि) — Climate-Smart Agronomy & Precision Irrigation Advisory
Research & Advisory System
Edge-AI & Precision Agronomy
Index 02Case Study

JalDrishti (जलदृष्टि)

Climate-Smart Agronomy & Precision Irrigation Advisory

State-of-the-art precision irrigation and crop advisory platform for Indian agriculture. Combines FAO-56 Penman-Monteith Evapotranspiration modeling, SoilGrids satellite physics, Smart Rain Hold warnings (≥5.0mm override), weather-driven pathogen risk forecasting for 5 critical crops, and JalSathi AI—a multilingual voice-enabled RAG agronomy assistant.

FastAPIPython 3.11FlutterPostgreSQLChromaDBHuggingFaceGroq Llama-3-70BOpen-Meteo APISoilGrids APIRedis Cloud
GitHub RepoResearch & Advisory Engine

Verified Honors & Academic Accreditations

Certifications & Credentials

Peer-reviewed presentation accreditations, hackathon awards, cloud infrastructure certifications, and industry engineering credentials.

Research & Presentation
Sep 202501

IEEE Conference Presentation Certificate

"An Explainable Deep Learning Approach for Quality Assessment in Solanaceous Crops"

Event: 2025 International Conference on Computing and Communications (COMPUTINGCON)

D Y Patil University, Ambi, Pune (IEEE Bombay Section)

Presented original research on explainable multi-task hybrid CNN-ViT models for simultaneous crop disease detection and commercial quality grading before an international peer panel.

Explainable AIComputer VisionConference PresentationResearch Publication
Hackathons & Competitions
Sep 202402

Smart India Hackathon — Internal Selection Winner

Event: Smart India Hackathon (Internal Selection Round)

The Neotia University

Selected as a winning team in the university-wide internal hackathon round for designing and presenting an innovative AI-powered solution for nationwide challenge problem statements.

Rapid PrototypingProblem SolvingSystem ArchitectureHackathon Defense
Academic Honors
202503

Certificate of Merit — Journal Publication

"AI-Driven Smart Waste Management System"

Anuranan (The Neotia University Interdisciplinary Journal)

Awarded academic merit certificate for authoring an interdisciplinary paper on computer vision and IoT sensor fusion for sustainable urban waste classification.

Academic WritingIoT & VisionWaste ClassificationEnvironmental AI
Advanced Training
Mar 202504

IEEE SPS Short Term Training Program (STTP)

Program: Revolutionizing Signal Processing: The Impact of AI and Machine Learning

Dream Institute of Technology & IEEE Signal Processing Society

Completed an intensive week-long professional training program on deep learning signal processing, biosignal analysis (ECG/sound), and neural feature extraction.

Signal ProcessingDeep LearningBiosignals (ECG)Feature Extraction

Academic Publications

Research & Publications

Peer-reviewed research in explainable deep learning, computer vision, and precision agricultural AI.

IEEE · Conference Proceedings18-19 September 2025Noida NCR, India

FruitQ-GradeX: Determining Fruit Quality and Grading with Explainable Deep Learning

Published in: 2025 12th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO)

Authors: Shibdas Dutta, Subhrendu Guha Neogi, Diya Chanda, Arpan Pramanik, Özgün Girgin, Enes Ladin Öncül

Key Results:98% fruit classification accuracy99% quality detection accuracy
IEEE Xplore
Springer · Springer LNNS21 July 2026Springer LNNS Book Series

Hyperspectral Fruit and Vegetable Classification Using Convolutional Neural Networks with EfficientNetB3

Published in: Data Mining and Information Security (ICDMIS 2025) — Lecture Notes in Networks and Systems (LNNS, volume 1915), Springer (pp. 449–469)

Authors: Shibdas Dutta, Subhrendu Guha Neogi, Arpan Pramanik, Diya Chanda, Özgün Girgin, Enes Ladin Öncül

Key Results:99.29% training accuracy97.21% test accuracyBenchmarked against 20+ architectures
SpringerLink
IEEE · Conference Proceedings18-19 September 2025Noida NCR, India

CropSense: Explainable Deep Learning Framework for Accurate Quality Detection in Solanaceous Crops

Published in: 2025 12th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO)

Authors: Shibdas Dutta, Subhrendu Guha Neogi, Shiladitya Chowdhury, Vikrant Chole, Arpan Pramanik, Diya Chanda

Key Results:99.9% crop classification accuracy98.5% quality detection accuracy30% lower parameter overhead
IEEE Xplore
IEEE · Conference Proceedings01-03 September 2025Talegaon, India

An Explainable Deep Learning Approach for Quality Assessment in Solanaceous Crops

Published in: 2025 International Conference on Computing and Communications (COMPUTINGCON)

Authors: Shibdas Dutta, Barshan Adhikari, Arpan Pramanik, Diya Chanda

Key Results:98.45% potato classification accuracy97.49% tomato classification accuracy98.5% quality assessment accuracy
IEEE Xplore

Get In Touch

Get In Touch

Feel free to reach out for research collaborations, engineering opportunities, or academic inquiries.

Direct Channels & Profiles

Emaildiya.chanda03@gmail.com
Location

West Bengal, India

LinkedIndiya-chanda2004
Connect
GitHubchandadiya2004
Repositories
ResearchGateDiya Chanda
Citations

Send a Message