Hi, I'm Kiarash.

Fifth-year student at Queen's University. I work on research and build things at the intersection of machine learning, systems, and real-world problems. Always looking for the next challenge.

GitHub LinkedIn

Contact

Awards

2nd Place, UofT ProjectX 2023
Shirley Bell Award 2025
2nd Place, Hack The Track 2024
Queen's Excellence Scholarship 2022
Certificate of Distinction, Fermat 2021
Leslie Wyber Scholarship (x2) 2021

Tools & Technologies

Python
PyTorch
TensorFlow
scikit-learn
NumPy
Pandas
LangChain
Pinecone
OpenCV
Hugging Face
React
TypeScript
Next.js
Flask
FastAPI
PostgreSQL

About Me

A bit about me

I'm a fifth-year student at Queen's University studying computing with a focus in artificial intelligence. I work on research, build systems, and enjoy diving deep into hard problems.

I'm always looking for interesting people to work with and challenging problems to tackle.

Education

Queen's University

BCmpH, Computing with Focus in Artificial Intelligence

2022 — Present

Involvement at Queen's

QMIND Director of Computer Vision · Competitive Research Team · AI Research Member
2022 - 2025
Queen's Startup ConsultingTechnical Consultant
2023 - 2024
Merlin NeurotechResearcher
2022 - 2023
inQubateVenture Founder
2024 - 2025
Queen's Space Engineering TeamElectrical and Software Member
2022 - 2023

Experience

Where I've worked

Jan 2025 — Present Kingston, ON

Founder & CTO at KynTrack

Founded and leading the technical development of KynTrack, a software platform.

May 2025 — Present San Francisco, CA

Data Science Consultant at PawCo Foods

Developing an internal chatbot using LLMs and RAG for pet nutrition queries, ingredient interactions, and product research.
Publishing the chatbot's architecture, domain-specific NLP design, and early testing insights.

Sep 2025 — Apr 2026 Kingston, ON

Teaching Assistant at Queen's University

Introduction to Cognitive Science (Jan 2026 - Apr 2026).
Fundamentals of Artificial Intelligence (Sep 2025 - Dec 2025).

May 2025 — Dec 2025 Guelph, ON

Research Assistant, AI for Computational Biology at University of Guelph

Led development of BeanGPT, the first domain-specific LLM for common bean genetics, integrating 750k+ papers, Pinecone, GPT-4, and NCBI. Published in Plant Methods.
Designed the full backend combining LLMs, vector search, and gene databases.

May 2024 — Aug 2024 Kitchener, ON

AI Engineer Intern at Amplify Care

Led fine-tuning of LLMs for a medical AI chatbot focused on billing inquiries, consulting 10+ physicians and clinicians.
Conducted cost-benefit analyses to optimize chatbot performance within budget.
Presented to senior managers, advancing the project to QA testing and deployment.

May 2023 — Aug 2023 Toronto, ON

Software Engineer Intern at BMO

Decreased debugging time by over 50% for job deployment via parallel processing.
Automated job calls to Ansible Tower using Python and cURL, presenting impact to senior managers.
Integrated microservices across teams, resolving complex technical challenges.

Publications

Research

Plant Methods · Jul 2026 First Author

BeanGPT: a domain-specific retrieval-augmented generation system for Phaseolus vulgaris research
↗

A domain-specific RAG platform for common bean research. Integrates scientific literature, agronomic trial data, climate projections, and genomic validation to generate accurate, citation-backed responses for researchers and plant breeders.
ProjectX 2023 · Jan 2024

Efficacy of K-Center Greedy Coresets: Streamlining Language Model Fine-Tuning for Abstractive Text Summarization
↗

Proposes using K-Center Greedy Coresets to select a representative subset of training data for LLM fine-tuning, achieving comparable performance on abstractive summarization with significantly less compute.
CUCAI · Apr 2023

EEG Brain-Computer Interface
↗

Using signal processing and advanced ML techniques to accurately detect blinks through brain signals, allowing users to play games hands-free via brain signal processing.

Projects

Things I've built

BeanGPT

↗

A domain-specific RAG system for Phaseolus vulgaris research. Integrates 750k+ papers, agronomic data, and genomic validation into a citation-backed assistant.

Python
RAG
Pinecone
GPT-4

EEG Brain-Computer Interface

↗

Signal processing and ML for hands-free gaming via blink detection. Our SVM hit 94.4% accuracy in 5 weeks. Showcased at CUCAI 2023.

Python
PyTorch
OpenBCI
Pygame

Touchless Browse

↗

A web browser requiring no mouse or keyboard. Navigate by voice via GPT-3.5 and Whisper, and simulate mouse movement using computer vision.

Python
OpenCV
Whisper
ChatGPT API

Flashcard Generator

↗

An NLP API leveraging BART and Sentence-BERT to automatically generate flashcards and precise answers from any text.

PyTorch
BART
Flask
Hugging Face

Awards

Recognition

2nd Place at UofT ProjectX 2023
2023

International ML research competition. Represented Queen's in a team of 4, placing 2nd among 20+ top global universities.
Shirley Bell Award

Recognizing academic excellence for students pursuing Mathematics, Statistics, or Mathematics and Engineering.
2nd Place at Hack The Track
Queen's University

Built a virtual autonomous racer that competed against 75 racers across 26 teams.
Queen's University Excellence Scholarship

Awarded for an admission average of 90 to 94.9%.
Certificate of Distinction, Fermat Contest
UWaterloo

Top 25% in the University of Waterloo's annual Fermat Mathematics Contest.
Leslie Wyber Excellence Scholarship (x2)

Awarded twice for exceptional musicianship and accelerated improvement.

Contact

Get in touch

I'm always open to new challenges, collaborations, and opportunities. Whether you have a question or just want to say hi, feel free to reach out.

kiarash.mirkamandari@queensu.ca LinkedIn GitHub

Hi, I'm Kiarash.

A bit about me

Where I've worked

Founder & CTO at KynTrack

Data Science Consultant at PawCo Foods

Teaching Assistant at Queen's University

Research Assistant, AI for Computational Biology at University of Guelph

AI Engineer Intern at Amplify Care

Software Engineer Intern at BMO

Research

BeanGPT: a domain-specific retrieval-augmented generation system for Phaseolus vulgaris research

Efficacy of K-Center Greedy Coresets: Streamlining Language Model Fine-Tuning for Abstractive Text Summarization

EEG Brain-Computer Interface

Things I've built

BeanGPT

EEG Brain-Computer Interface

Touchless Browse

Flashcard Generator

Recognition

2nd Place at UofT ProjectX 2023

Shirley Bell Award

2nd Place at Hack The Track

Queen's University Excellence Scholarship

Certificate of Distinction, Fermat Contest

Leslie Wyber Excellence Scholarship (x2)

Get in touch