Skip to content
View kkirpa's full-sized avatar

Block or report kkirpa

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kkirpa/README.md

👋 Hey, I'm Kirpa!

I find patterns, build solutions, and chase impact through data.

I'm obsessed with connecting the dots between raw data and real-world change, AND learning something new every time I do it.


What Drives Me

  • Digging into complex data and finding meaning
  • Building usable, scalable tools, not just shiny models
  • Tackling problems that matter to real people
  • Picking up new tech like it’s a puzzle I have to solve

My goal is to bridge the gap between cutting-edge Technologies and the humans who actually use it. Whether it's translating the “black box” of AI agents into something stakeholders understand, or building interfaces that connect complex models to real-world impact, I bring the tech down to earth and make it make sense.


Some of my Project Work:

Facial Recognition System

Real-time facial ID system using ResNet50 + Attention + Triplet Loss (~95% accuracy). → Built FastAPI + React interface with 1:1 and 1:N matching → Roadmap: Improve fairness, speed, and deploy liveness detection

Alzheimer’s Early Detection

Built 95% accurate models using clinical data from 2,149 patients.
→ Prioritized interpretability and bias checks.
→ Exploring deployment as a clinical support tool.

Pharma ML at Bristol Myers Squibb

Built real-time ML pipeline to replace expensive CFD simulations.
→ Developed FastAPI app for both technical + business users.
→ Focused on scalability, stakeholder alignment, and future monitoring.

NYC Pedestrian Safety Analytics

Mapped 3 years of crash data to spot danger zones in NYC.
→ Used geospatial tools to recommend Vision Zero policy changes.
→ Planning real-time dashboards for urban planning.

X-Scape: Accessible VR Escape Room

Led development of a multiplayer escape room for all abilities.
→ Combined Unity, haptics, and spatial UX design.
→ Won 2nd place at VT HCI Symposium 🥈


Skills and Tech Stack

Languages & Libraries:
Python (pandas, NumPy, scikit-learn, PyTorch), SQL, R, JavaScript

Frameworks & APIs:
FastAPI, Flask, React.js, REST

Visualization & Storytelling:
Tableau, Power BI, matplotlib, seaborn, ggplot2, GeoPandas

Dev & Cloud:
Docker, Git/GitHub, AWS (S3, Lambda), PostgreSQL

Soft Skills:
Communication, DEI-aligned leadership, cross-team coordination, conflict resolution


My future projects

I’m constantly leveling up — right now I’m diving into:

  • Real-time ML monitoring + MLOps
  • Explainability + fairness in predictive systems
  • Agent-agents for Mental Health Support
  • Interactive dashboards that actually make sense, and can update real-time

What I’m Looking For

I’m exploring full-time roles in:

  • Data Science, Analytics / ML Engineering
  • Applied Analytics & Insight Strategy
  • Solutions Architect / Data Consulting
  • Accessible Software/Application Development

I want to work on stuff that matters , with teams that think critically, build thoughtfully, and ship responsibly.


A Few Fun Facts

  • I unwind with music and cultural dance
  • I used to do competitive shooting (yeah, really)
  • You’ll find me hiking and reflecting when I'm not buried in code
  • I lead data workshops and advocate for DEI in tech

Connect with me

Reach out if you’re hiring, vibing with my work, or want to nerd out over new technology developments, pleade feel free to reach out via:

💼 LinkedIn🐙 GitHub📫 Email


Thanks for stopping by 💫
– Kirpa

Pinned Loading

  1. Face-Recognition-RAG-Sytem Face-Recognition-RAG-Sytem Public

    Forked from JavierAM01/Face-Recognition-RAG-Sytem

    A FastAPI web app for face verification (Siamese ResNet) and face search (InceptionResnetV1) with live webcam registration.

    Python

  2. alekhyavittalam/alzheimers_diagnosis_prediction alekhyavittalam/alzheimers_diagnosis_prediction Public

    Leveraging machine learning to develop a predictive model for early and accurate Alzheimer’s disease diagnosis. The aim is to guide test prioritization and improve diagnostic efficiency, to support…

    Jupyter Notebook 1

  3. ComputerVision ComputerVision Public

    Portfolio of CMU computer vision projects showcasing implementations of numerous foundational topics

    Jupyter Notebook

  4. Diamonds_Price_Prediction--ML Diamonds_Price_Prediction--ML Public

    In this study, our team explored and applied a range of machine learning models to predict diamond prices based on key features and characteristics of the diamonds.

    R 1

  5. Large-Scale-Computing Large-Scale-Computing Public

    Large Scale Computing projects done at CMU, as a Data Science Master's student. Done on Bridges-2 Supercomputer at Pittsburgh Supercomputing Center.

    Python

  6. X-Scape--VR_Escape_Room X-Scape--VR_Escape_Room Public

    Project "X-Scape" is an Escape Room in VR, can also be played on PC by building the app for Windows/Mac from Unity. [developed in Unity for HCI Capstone 2024, Virginia Tech]

    C# 3