Dhanush N - Profile Picture

AI SYSTEMS BUILDER | SOFTWARE DEVELOPER | ML & AGENTIC AI

Building Intelligence
From Data to Deployment

I build

AI & Data Science student who builds LLM-powered tools — from RAG pipelines to voice assistants to 3D molecule learning tools. National hackathon champion with a track record of turning ideas into working prototypes fast. Currently a Software Developer Intern at Adya AI, shipping features to a live production platform and building AI data-analyst tooling.

View Projects Contact Me
3Hackathon Wins
6Competitions & Ranks
8.72CGPA

About Me

I am a B.Tech student in Artificial Intelligence and Data Science at Karpagam College of Engineering.

I build end-to-end AI and full-stack applications, working across machine learning, computer vision, and backend development.

I focus on writing clean, scalable code and developing systems that solve real-world problems effectively.

Since May 2026 I have been a Software Developer Intern at Adya AI, combining LLM reasoning, retrieval-augmented generation, analytics, automation and full-stack engineering — from client defect resolution and feature delivery to FastAPI services, conversational data assistants and interactive 3D visualization.

Education

B.Tech AI & Data Science

Karpagam College of Engineering | 2023 - Present

CGPA: 8.72 / 10

HSC (Higher Secondary)

Sree Vidhya Vikas Matric School | 2022 - 2023

Score: 94%

Experience

Software Developer Intern

Adya AI

May 2026 – Present Remote Current role

Castlewise

Client Project

Role-based, multilevel platform for housing society and apartment redevelopment (live in production).

Live in productionRole-based accessJira & QA
  • Tender & Business FlowImplemented the end-to-end tender flow and core business workflows, covering the tender lifecycle for society and apartment redevelopment projects.
  • Multilevel Role-Based SystemBuilt role-based access controls for screens, permissions and actions so each stakeholder sees functionality relevant to their role.
  • Platform ModulesImplemented the Notification module, Payment Master and super-admin screens for user alerts, centralized administration and payment configuration.
  • Feature DeliveryDelivered client-requested features based on feedback while keeping changes aligned with agreed requirements.
  • Defect Resolution & QAResolved Jira-tracked defects, verified fixes through testing, maintained ticket updates and covered QA responsibilities during tester absence.
  • Client CollaborationJoined client meetings to clarify requirements and fix/feature questions, and shared ticket-status updates to support coordination between client and team.

RAS

Product Development

AI data-analyst product: forecasting, 3D visualization and narrated story presentations.

Data Analyst agent3D visualizationForecasting
  • UI ModernizationRedesigned and refreshed the user interface to improve usability and visual consistency.
  • ForecastingFine-tuned the Data Analyst agent's forecasting capability to improve forecasts generated from user datasets.
  • 3D Chart GenerationExtended the Data Analyst agent with 3D chart generation so results can be presented as 3D visualizations instead of standard 2D charts.
  • 3D DashboardBuilt a unified 3D dashboard combining globe, bar and line charts in one visual workspace.
  • AI Story ModeDeveloped a story mode that converts analysis results into animated slide presentations with AI voice narration.
  • Platform ModulesImplemented Notification, Payment and Sharing master modules for alerts, payment handling and content sharing.

Technical Skills

Languages

PythonJavaGoJavaScriptSQL

AI / ML

Machine LearningData Science Computer VisionNLPDeep Learning OpenCVYOLO LLMsAgentic AIRAGSemantic Search LightGBMXGBoostScikit-learn TensorFlowKeras

Web & Backend

FastAPIFlaskStreamlit Spring BootREST APIs

DBs & Tools

ChromaDBPostgreSQLMySQLSnowflake KafkaDockerMLOpsGit SQLitePineconeFAISS Docker ComposeJiraJupyter NotebookVS Code

Data & Streaming

PandasNumPyApache KafkaPathway Data AnalysisAnomaly DetectionForecasting

Visualization & 3D

Chart.jsTableauPower BIOpenCV BlenderUnity3DARCore/ARKit

Projects

📄

TalentAI

Resume Intelligence Platform

AI-assisted recruitment workflow that scores resumes against job descriptions, extracts candidate details and ranks profiles by semantic relevance.

PythonFastAPILLMsRAGNLPChromaDBMySQLJavaScript

Click to view features

Key Features

  • Resume intelligence and job-description matching
  • Retrieval-augmented ranking with FastAPI, RAG and ChromaDB
  • GitHub and LinkedIn profile checks in the review workflow
View Code
🎙️

Flexie 2.0

Voice-Controlled Agentic Desktop Assistant

Windows-native assistant that pairs voice interaction with desktop and web actions for hands-free task execution.

PythonFastAPITool OrchestrationMemoryDesktop Automation

Click to view features

Key Features

  • Hands-free voice and desktop automation
  • Goal planning, tool registration and DAG-based task planning
  • Workflow execution with automation verification
  • Memory and tool orchestration for multi-step tasks
View Repository
🎓

TNEA Pro AI

Admissions Counselling Assistant

Answers questions from TNEA brochures and policy documents by combining retrieved context with structured cutoff data.

FastAPIGroq/LlamaOllamaChromaDBSQLiteLangChainDocker

Click to view features

Key Features

  • RAG-based counselling over brochures and policies
  • Safe, Moderate and Dream recommendations from 2021–2025 cutoffs
  • Searchable college profiles, choice-list exports and TFC centre locator
  • Query caching, debouncing/aborting and Docker Compose support
View Repository
🔬

3D Molecule Visualizer

AI-Based Blender Pipeline

Converts molecular XYZ coordinate data into 3D structures, with a lightweight web interface for exploring them.

PythonBlender Python APIHTML/CSS3D Model Export

Click to view features

Key Features

  • XYZ data to 3D scenes and models via Blender scripting
  • Reusable GLB exports with a lightweight web interface
  • Voice-assisted interaction for science learning
View Repository
♻️

E-Waste Deconstructor

Computer Vision & Digital Twin

YOLO-based workflow that identifies and classifies e-waste items and estimates hazard severity to inform safer handling.

PythonYOLOOpenCVSimulationHTML/CSS

Click to view features

Key Features

  • Classification of e-waste items with hazard severity estimates
  • Digital-twin simulation with a live view of process stages
  • Safe-deconstruction recommendations on a web dashboard
View Repository

GitHub

-

Public Repositories

-

Total Stars

-

Total Forks

Loading projects from GitHub...

Coding Profile

LeetCode Stats

-

Problems Solved

-

Total Submissions

-

Global Ranking

Easy

-

- total

Medium

-

- total

Hard

-

- total

View Full Profile

Achievements

02/2026

🏆 Top 10 - Hacknation 2026

4th Global AI Hackathon.

01/2026

🥇 National Hackathon Winner

IIT Kharagpur & Shiv Nadar University.

10/2025

🥇 AIM'25 Winner

Secured first place at Infynd's Hackathon.

09/2025

🥇 HackZilla 25 Winner

KPR Institute of Technology, IEEE ComSoc.

09/2025

🎯 Top Finalist - HackBattle'25

VIT Vellore.

03/2025

🏅 33rd Rank - FIFS Gameathon 2.0

National-level sports data analytics competition.

Certificates

Education for Sustainable Development

NPTEL (Elite + Silver, 86%)

Introduction to IoT

NPTEL (Elite + Silver, 76%)

Python for Data Science

NPTEL (Elite, 74%)

Enhancing Soft Skills

NPTEL (Elite, 69%)

Get In Touch

I am currently open to internships and collaborative AI projects. Let's build something amazing together!

Based in Salem, Tamil Nadu · currently a Software Developer Intern at Adya AI.

Email LinkedIn GitHub Resume (Drive) +91 9342514011