Skip to content
HomeCommunityProjectsExperienceHackathonsContact
Portrait of Ishneet Kaur Chadha
  • Drawn to problems that reward patience and precision.
  • Enjoys thoughtful discussions and careful iteration.
View ResearchView ProjectsDownload Resume
LinkedInGitHubGoogle Scholar

University of Colorado Boulder

Ishneet Kaur Chadha

"I enjoy building software that’s easy to read, easy to reason about, and built to last. My work focuses on clean, maintainable code, collaborative problem-solving, and adapting quickly as ideas evolve. As a Computer Science graduate student working across systems, machine learning, and research-oriented projects, I care as much about clarity and design as I do about results and I’m always happy to connect and talk through interesting problems."

Machine Learning & NLPSystems & InfrastructureResponsible AI
ishneet.chadha@colorado.edu

Current focus

I’m working on questions around state, consistency, and evaluation in learning driven pipelines, especially in settings where data arrives incrementally rather than as a static dataset.

Areas of attention

  • Evaluation and benchmarking for NLP systems
  • Reproducible ML experimentation pipelines
  • Bias, data quality, and model reliability

Research

Research spotlight

Three active areas where I am building evaluation-first, trustworthy ML systems.

Performance Evaluation of Open-Source LLMs for Text-to-SQL Conversion in Healthcare

First-author study presented at ICPCT 2025 and published on IEEE Xplore, evaluating open-source LLMs (Llama 3.1, Mixtral, Gemma 2) on MIMICSQL for clinically grounded Text-to-SQL translation. The evaluation emphasizes empirical rigor through exact-match accuracy and logical-form correctness, revealing clear scale–reliability trade-offs and positioning Llama 3.1–70B as the most dependable model for high-stakes clinical decision support.

Healthcare NLPOpen-source LLMsClinical decision supportMIMICSQL benchmarkExact-match accuracyLogical-form correctness
Read paper

Cloud Computing Applications in Digital Health: Challenges Related to Privacy and Security

Co-authored the Springer edited book chapter “Cloud Computing Applications in Digital Health: Challenges Related to Privacy and Security” in Explainable IoT Applications: A Demystification (Information Systems Engineering and Management, vol. 21) under Dr. Tripti Rathee. The chapter examines cloud service models, privacy and safety threats, and security methods for digital health systems, with real-world applications such as telemedicine, medical imaging, clinical information systems, and IoMT. First online: 14 Feb 2025 (pp. 79–97).

Cloud computingIoT applicationsDigital healthPrivacy & securityEthicsGovernanceApplied ML
View chapter

SemEval-2025 Task 6: Clarity & Evasion Detection (QEvasion / CLARITY)

Study on detecting strategic ambiguity in political interviews by classifying question–answer pairs into Clear Reply, Ambivalent Reply, and Clear Non-Reply. Using the QEvasion dataset of U.S. presidential interviews, we conduct EDA on label distributions and answer-length patterns, implement BERT-based baselines, and evaluate macro F1 while analyzing error sources and limitations.

Political NLPQuestion–answer classificationBERT baselineMacro F1Exploratory data analysisEvasion detection

Work in progress

Hackathons

Selected hackathon wins

Highlights from rapid-build competitions focused on applied AI, systems, and sustainability.

Cal Hacks × Fetch.ai Innovation Lab

Winner, Fetch.ai track — AgentViewAR

Built an empathy-driven VR interview coaching system that surfaces real-time stress insights to recruiters. A Meta Quest 3 HUD shows pace, pauses, tension levels, and prompts; Fetch.ai agents coordinate recruiter/candidate feedback loops. Groq Whisper handles fast speech transcription, FastAPI + ChromaDB track recovery patterns, and Claude produces interview summaries and coaching recommendations.

Meta Quest 3WebXRThree.jsVR HUDWebSocketsFastAPIWhisperASRSpeech analyticsClaude APILLM summarizationFetch.ai AgentverseMulti-agent systemChromaDBVector database
DevpostMedium articleGitHub

HackHazards ’24

Winner, xxnetwork track — CrimsonConnect

Built a transparent, user-friendly platform that connects donors and recipients quickly, improving timeliness while preserving privacy with blockchain-based verification.

Implemented a Django web app (auth, templates, static assets) and integrated Solidity/Hardhat with Ethereum tooling and Polygon Mumbai. Added location capabilities with Leaflet.js and navigated xxNetwork ecosystem constraints and test token limitations.

SolidityNode.jsDjangoJavaScriptReact.jsHardhatEthereum TestRPCPolygon MumbaiLeaflet.jsxx-network
DevfolioYouTube DemoGitHub

GeeksforGeeks Ecotech Hacks

2nd place — OptiSpace

OptiSpace is a smart office system that combines IoT sensing, machine learning, and computer vision to optimize energy usage, occupancy-aware climate control, and waste management. It ingests temperature, humidity, lighting, and door-entry signals alongside OpenWeatherMap context, then applies predictive + rule-augmented ML to adjust lighting and HVAC. A lightweight vision pipeline detects waste and feeds a dashboard for real-time monitoring and energy insights.

IoT & sensingReal-time inferenceMachine learningPredictive modelingDjangoRESTful APIsBackend orchestrationComputer vision
YouTube DemoGitHub

AWS GameDay Hackathon

2nd place — team performance highlight

Secured 2nd place while collaborating under time pressure using Amazon Bedrock, OpenSearch, SageMaker, and AWS Guardrails. This was a competitive team challenge rather than a single product build.

Amazon BedrockOpenSearchSageMakerAWS GuardrailsCloud architectureGenAI workflows

Timeline

Education and internship timeline

A snapshot of academic milestones and applied industry experience.

Education

Internships

2025

Aug 2025 – Present

Master of Science in Computer Science

University of Colorado Boulder

Jan 2025 – Jun 2025

Data Analyst Intern, Decision Tree Analytics

Gurugram

2024

Aug 2024 – Nov 2024

AI/ML Intern, TEMS Tech Solutions

Remote

Jul 2024 – Aug 2024

AI/ML Intern, Tech Mahindra

Remote

2023

Jul 2023 – Sep 2023

Research Trainee, Defence Research and Development Organization

New Delhi

2021

Aug 2021 – Jun 2025

Bachelor of Information Technology

Undergraduate program