Portfolio ©2026Bengaluru, IN
Models
0
Open to opportunitiesBengaluru, India

I spend my days orchestrating AI agents

I build AI-powered products and the backends that keep them running, from LLM agent workflows and computer vision pipelines to real-time, event-driven systems.

Python✦PyTorch✦LangGraph✦FastAPI✦Django✦React✦TensorFlow✦Kafka✦Docker✦AWS✦Python✦PyTorch✦LangGraph✦FastAPI✦Django✦React✦TensorFlow✦Kafka✦Docker✦AWS✦
Multi-Agent Systems✦Computer Vision✦Real-Time Backends✦LLM Workflows✦Event-Driven Systems✦REST APIs✦Multi-Agent Systems✦Computer Vision✦Real-Time Backends✦LLM Workflows✦Event-Driven Systems✦REST APIs✦

ComputerSciencegraduatewholikesshippingthingsthatactuallyworkinproduction.I'vebuiltmulti-agentsystemsthattailorresumeswithoutinventingasingledetail,reconstructed3Dpointcloudsfromasinglesilhouette,trainedCNNsthatexplaintheirownplant-diseasediagnoses,andengineeredaproctoringbackendthatscorescheatingriskfromliveexameventsinrealtime.IcareaboutcleanAPIs,honestAI,andinterfacespeopleenjoyusing.

0

AI & full-stack projects built and shipped

0

User roles secured with JWT-based RBAC at PyGenicArc

0.00

CGPA in B.E. Computer Science

0

Certifications from Stanford, DeepLearning.AI, AWS & Microsoft

Where I'veshipped real systems.

Feb 2026 – May 2026

Bengaluru

Python Backend Developer Intern

@ PyGenicArc

Built the backend for a remote proctoring platform with real-time monitoring and anti-cheating.

  • 01Built the backend of a remote proctoring platform with Django REST Framework, PostgreSQL, Redis, Kafka, WebSockets, LiveKit and Docker Compose, with JWT-based RBAC for 4 user roles.
  • 02Self-hosted LiveKit via Docker and built WebSocket session management with live stream tracking, heartbeats and reconnection handling.
  • 03Implemented an anti-cheating engine with automated risk scoring, violation tracking (tab switching, focus loss, multi-person detection), Kafka event streaming and exam termination logic.
  • 04Built the full exam lifecycle API suite: question banks, scheduling, batch management with Excel upload, attempt tracking, automated evaluation, results and analytics dashboards.
Django RESTPostgreSQLRedisKafkaWebSocketsLiveKitDocker

Things I'vebuilt & shipped.

Full-Stack
012026

CareerAtlas

Multi-agent AI job search platform

Upload a PDF resume, pick a target role, and CareerAtlas finds live listings, ranks them against what your resume actually shows, and generates a skill-gap analysis, tailored resume and cover letter.

ReactFastAPILangGraphGeminiSupabaseVercelRender
View details
AI / ML
022026

recon3d

One silhouette in, a 3D point cloud out

A ResNet-18 encoder and MLP decoder predict 2,048 3D points from a single silhouette image, trained end to end with Chamfer distance on six ShapeNet categories and served as a FastAPI + Gradio app that warns you when an upload looks unlike anything it was trained on.

PyTorchResNet-18FastAPIGradio
View details
AI / ML
032026

OMNIMESH AI

Sketch-to-3D mesh generation

Draw or upload a 2D sketch, describe it in a prompt, and get back textured 3D variations you can orbit, colorize, edit, view in AR/VR and export as .obj or .stl.

PyTorchControlNetTripoSRThree.js
View details
AI / ML
042025

Resume Analyzer Pro

ML + NLP resume screening

Reads a resume, predicts the best-fit job category, extracts key skills, scores the resume and gives actionable feedback, for students and hiring managers alike.

PythonScikit-LearnNLPStreamlit
View details
AI / ML
052026

Potato Leaf Disease Detection

CNN diagnosis with Grad-CAM explanations

A convolutional neural network that tells early blight, late blight and healthy potato leaves apart — and shows a Grad-CAM heatmap of the lesions it actually looked at, so a diagnosis can be trusted.

TensorFlowKerasCNNGrad-CAM
View details
AI / ML
062026

Housing Price Prediction

Linear regression written from scratch

Multiple linear regression built from scratch with NumPy — no scikit-learn — to estimate house prices, with the trained parameters served through a Streamlit app.

PythonNumPyPandasMatplotlib
View details
AI / ML
072026

Diabetes Prediction

SVM clinical risk classifier

A support vector machine that predicts diabetes risk from eight diagnostic measurements, served through a Streamlit app that shows its own accuracy alongside every prediction.

PythonScikit-LearnSVMPandas
View details
Full-Stack
082025

Olympics Medal Tracker

Paris 2024 medal dashboard

A web app to track Paris 2024 medal counts by country, with charts, search, event highlight videos and full medalist management.

FlaskJinja2JavaScriptSQL
View details

My everydaytoolkit.

01

AI / ML

◎
  • PyTorch
  • TensorFlow
  • Scikit-Learn
  • NumPy
  • LangGraph
  • Gemini
  • CNN
  • Grad-CAM
  • RNN / LSTM

02

Backend

⌘
  • Python
  • FastAPI
  • Django REST
  • Flask
  • PostgreSQL
  • Redis
  • Kafka
  • WebSockets

03

Frontend

◐
  • React
  • TypeScript
  • JavaScript
  • HTML
  • CSS
  • Tailwind CSS

04

Cloud & Tools

☁
  • Docker
  • AWS
  • Hugging Face
  • Supabase
  • Vercel
  • Git
  • GitHub
  • LiveKit
  • Streamlit
  • Gradio

Alwayslearning.

2022 – 2026Bengaluru

K S School of Engineering and Management

B.E. in Computer Science and Engineering

8.04 CGPA

2020 – 2022Bengaluru

Narayana Pre University College

PUC (PCMC)

84.5%

  • Deep Learning SpecializationDeepLearning.AI & Stanford University
  • Machine Learning SpecializationDeepLearning.AI & Stanford University
  • AWS Academy Graduate – Cloud DevelopingAmazon Web Services
  • Applied Artificial Intelligence: Practical ImplementationsTechSaksham by Microsoft & SAP · Edunet Foundation · AICTE