K
KGEN
Building and maintaining full stack web applications, working across both frontend and backend to ship and iterate on production features.
I'm Aditi Singh.
Software Engineering Student. Passionate about creating, innovating, and solving challenges through code.
About
Currently pursuing a B.Tech in Computer Engineering, I’m passionate about building things and solving challenging problems. Over the past few years, I’ve developed multiple side projects, competed in competitive programming contests, and explored emerging technologies. I enjoy finding creative solutions, both in code and in real-world applications, and learning through hands-on experiences.
Work Experience
Skills
React
Python
Java
C++
JavaScript
TypeScript
MatPlotLib
Panda
NumPy
Git
Docker
Scapy
PyShark
Streamlit
TensorFlow
Librosa
Scikit-learn
Convolutional Neural Networks (CNN)
Figma
Adobe Photoshop
Adobe Premier Pro
Adobe AfterEffects
Adobe Illustrator
Activity
My Projects
Here's what I've been up to!
I've worked on a variety of projects, from simple websites to complex web applications. Here are a few of my favorites.
SAMHITA - Agentic Compliance Platform
An agentic regulatory-compliance platform built on a five-agent LangGraph pipeline, with XGBoost routing at 94% accuracy and a SHA-256 audit ledger for RBI/SEBI obligations. The explainable compliance engine pairs a deontic-logic parser (0.94 F1) with UMAP visualization and Monte Carlo risk forecasting.
Python
FastAPI
Next.js
PostgreSQL
LangGraph
XGBoost
UMAP
Meridian
An AI-powered intelligence and observability platform that aggregates organizational data from GitHub, Jira, Slack, and OpenMetadata to deliver role-specific insights for executives, project managers, HR, and finance teams. Built as a multi-agent LLM platform with real-time dashboards, automated standups, and predictive analytics.
FastAPI
Next.js
React
TypeScript
GitHub API
OpenMetadata
WebSockets
Disha: Urban Traffic Intelligence
A GCN-LSTM framework for traffic forecasting that models spatial road dependencies and temporal congestion patterns from real-world transportation data. Evaluated against statistical and persistence-based baselines across varying traffic conditions to improve congestion prediction and urban mobility planning.
Python
PyTorch
Graph Convolutional Networks (GCN)
LSTM
Time Series Forecasting
Advanced Prenatal Detection of Fetal Neurological Abnormalities
Non-invasive prenatal detection of fetal neurological abnormalities using CycleGAN-based fetal ECG reconstruction and heart rate variability (HRV) analysis from maternal abdominal ECG recordings. Ant Colony Optimization drives feature selection and Grey Wolf Optimization tunes the Random Forests, extracting 12 clinically validated HRV biomarkers validated through Leave-One-Subject-Out Cross Validation across 80 subjects.
Python
CycleGAN
Random Forest
Ant Colony Optimization
Grey Wolf Optimization
Honours
Pushing Limits, Earning Laurels
Recognized for excellence in coding, innovation, and teamwork, these honors highlight my journey of perseverance, creativity, and impact.
- W
Winner of Meesho Scripted by Her 2.0
Meesho
Winner of Meesho Scripted By Her 2.0 for developing KrishiMitra, an AI-powered agricultural assistant leveraging a multi-agent architecture to provide farmers with crop health diagnosis, weather insights, market intelligence, financial guidance, and risk assessment. - W
Winner of SuRaksha Cyber Hackathon 2.0
Canara Bank
Won the SuRaksha Cyber Hackathon 2.0, hosted by Canara Bank, for developing an AI-driven solution for detecting spurious gold and strengthening the security and reliability of gold-based lending. - W
Winner of Project Q2 Research Hackathon
Project Q2
Winner of the Project Q2 Research Hackathon for proposing an AI-assisted framework for prenatal neurological risk screening in low-resource healthcare settings, focusing on accessible, early-stage risk assessment through AI-driven research and innovation. - P
Pupil @ Codeforces
Codeforces
Reached Pupil rating on Codeforces through consistent participation in competitive programming contests. - W
Winner of Coderush 5.0: The IVP Hackathon
Indus Valley Partners
Secured the top position out of 1,100+ candidates across Computer Science, DSA, and Web Development rounds. - R
Runner-up - CodeCraft CTF '25
CodeCraft
Cybersecurity Capture The Flag competition focused on solving security challenges across multiple domains. - W
Winner of Code Without Barriers Hackathon
Microsoft
Participated in the Code Without Barriers Hackathon, made an intelligent PCAP Analysis tool for Trading Systems, Exchlytics AI. - 2
2 Star @ CodeChef
Codechef
Achieved 2-star rating on CodeChef, showcasing strong problem-solving skills and consistency in competitive programming. - S
SIH 2024 Finalist
Chennai, India
Designed and developed a Drug Inventory and Supply Chain Tracking System, streamlining medication distribution with real-time tracking, ensuring efficient stock management, and minimizing wastage. Focused on enhancing transparency and accessibility for underserved populations. - C
Contributor @ Kaggle
Kaggle
Active contributor on Kaggle, leveraging data analysis and machine learning to solve real-world problems and participate in global competitions. - S
SheFi Scholars Program
Remote
Received the SheFi Scholars Program scholarship, which provided comprehensive training in decentralized finance (DeFi) and blockchain technology, enhancing my understanding of the crypto ecosystem and its applications.
Open Source
Building in the Open
Projects I follow and contribute to: agent governance, AutoML, confidential databases, and internet measurement.
Forked from microsoft/agent-governance-toolkit
AI Agent Governance Toolkit: policy enforcement, zero-trust identity, execution sandboxing, and reliability engineering for autonomous AI agents. Covers 10/10 OWASP Agentic Top 10.
Python6,272
Forked from microsoft/FLAML
A fast library for AutoML and tuning, covering hyperparameter search and cost-effective model selection.
Jupyter Notebook4,395
Forked from InternetHealthReport/ihr-website
Vue.js front end for the Internet Health Report, visualising global internet measurement data.
Vue93
Forked from InternetHealthReport/internet-yellow-pages
A knowledge graph for the Internet, linking networks, prefixes, and routing data into one queryable graph.
Python82
Forked from agentrust-io/agent-manifest
Agent Manifest SDK hardware-anchors all 10 artifacts that define an agent at deployment. Python and TypeScript.
Python19
Forked from cyborginc/cyborgdb-py
Python client SDK for CyborgDB, a confidential vector database for encrypted embedding search.
Python8
Licenses & Certifications
Always Learning Something New
- D
Structuring Machine Learning Projects
DeepLearning.AI
Credential ID 955a9c41-3f82-4f6c-9548-5d2c45b7e863
ML System DesignPerformance Tuning - D
Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
DeepLearning.AI
Credential ID e238c1b5-fc4e-4bb5-a92d-d80b50724fba
Hyperparameter TuningModel Optimization - D
Neural Networks and Deep Learning
DeepLearning.AI
Credential ID d8e35697-86ab-4677-a8d2-2748597432d5
Neural NetworksConvolutional Neural Networks (CNN) - G
Google Data Analytics
Google
Credential ID 2G7TZXILKEX0
Data AnalysisData Cleaning