Associate Software Engineer · AI / ML Researcher
Building intelligent systems at the intersection of software engineering, machine learning, and enterprise technology. Based in Chandigarh, India.
About Me
I'm a Computer Science graduate from CCET Chandigarh with a strong foundation in AI/ML, full-stack development, and enterprise software. I am driven by a deep curiosity for how intelligent systems are designed and built.
Currently at rtCamp as an Associate Software Engineer, I work on Frappe-based internal tools and client-facing features. Before that, I engineered 30+ ERP modules at KoreCent Solutions — independently owning full development cycles from requirements to deployment.
My research spans computer vision, NLP, and clinical AI — published in ACM, Springer, and a Q1 journal (Impact Factor: 2.7). I care about building things that actually work in the real world.
What I Work With
Career
Things I've Built
Frappe-based CLM with role-based access, GenAI contract summaries, Documenso e-signing, 10+ stage approval workflows, versioning, audit readiness, and real-time notifications via Socket.IO + Redis.
AI-powered research assistant integrating academic APIs, GenAI literature review tables, and multilingual translation via Sarvam AI. Reduced literature review turnaround to 15–25 seconds.
Flask web app for car price prediction using Gradient Boosting Regression (R² = 0.95) with SelectKBest feature selection and real-time user inputs.
Full-stack app (Flask, PostgreSQL, HTML/CSS/JS) with REST APIs, book metadata integration, e-invoice ARN generation, CRUD operations, issue/return workflows, and real-time dashboards.
Fine-tuned Meta Llama 2 (7B) on domain-specific datasets using Amazon SageMaker, enhancing text generation performance on specialized topics.
Streamlit app using a LoRA-fine-tuned BART LLM for abstractive summarization and Indian language translation via Sarvam AI API.
User-centered mobile app prototype for a local bakery. Conducted user research, built personas, mapped user journeys, and created an interactive high-fidelity Figma prototype.
CNN-based gesture recognition using PyTorch and OpenCV, achieving 100% accuracy on a custom dataset of 1,500 images across multiple gesture classes.
Academic Work
Proposed a custom 3-tier CNN (VGD6-NET) achieving 98.5% accuracy across 6 waste classes for automated waste segregation.
Explored advanced neural networks for robust fake news detection across heterogeneous corpora.
Improved CVD risk prediction accuracy from 0.83 → 0.91 using ML on hybrid clinical datasets.
Advanced high-risk Sepsis-III patient care through deep learning-based mortality prediction.
Recognition
Selected among top 20 from 500+ applicants. Trained in Image Processing, Computer Graphics, and Computer Vision.
Verify Credentials ↗Ranked Top 12 out of 250+ students (from 50,000+) at Hitbullseye's national competition.
Verify Credentials ↗Sponsored by AWS to pursue Udacity's AI Programming with Python Nanodegree.
Verify Credentials ↗Won second prize at the Innovative Product Expo, Chandigarh College of Engineering and Technology.
Verify Credentials ↗Secured second place in the Botpull inter-college robotics competition.
Verify Credentials ↗Delivered a lecture on Robotics & Embedded Systems to 50+ high school students. Received an official Appreciation Letter.
Verify Credentials ↗Learning
Get In Touch
Whether it's a project, a collaboration, a research opportunity, or just a conversation — I'd love to hear from you.