Available to hire
Hi, I’m Hrayr Muradyan, a passionate data scientist and machine learning engineer based in Vancouver, Canada. I love turning complex AI concepts into accessible knowledge, which I share through my AI-focused YouTube channel called AI For Beginners. Balancing full-time studies with remote work, I’ve gained valuable experience working on cutting-edge projects in companies like ServiceTitan and Philip Morris International.
My expertise spans natural language processing, computer vision, and deep learning, combined with strong skills in Python, AWS, and data visualization. I’m always excited about leveraging AI to solve real-world problems and am eager to continue growing in this dynamic field.
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Language
English
Fluent
Work Experience
Machine Learning Researcher at ServiceTitan
December 31, 2025 - July 18, 2025Implemented Retrieval-Augmented Generation systems with Nearest Neighbours Search to improve sales proposal generation. Reduced custom evaluation metric's median score from 11 to 4, enhancing proposal quality. Fine-tuned BERT to predict optimal proposal length, reducing baseline mean absolute error by about 40%. These improvements are expected to boost proposal sales by approximately 4%, potentially generating an estimated $10M additional revenue.
Data Scientist at HCL Technologies
December 31, 2025 - July 18, 2025Developed an end-to-end Retrieval-Augmented Generation (RAG) system for automated incident resolution with a reproducible Makefile setup, tests, and a Quarto report. Evaluated Google and OpenAI embedding models using RAGAS, achieving over 78% faithfulness and 62% correctness, demonstrating promising results for large scale evaluation. Conducted LLM-as-a-judge evaluations across resolution strategies; ANOVA confirmed significant performance improvements for RAG systems (p < 0.01). Tested embedding robustness to character-level noise, finding approximately 4% metric drop per 10% corruption.
Machine Learning Researcher at ServiceTitan
December 31, 2025 - July 18, 2025Implemented RAG systems with Nearest Neighbour Search to improve sales proposals quality, reducing median custom evaluation score from 11 to 4. Fine-tuned BERT to predict optimal proposal length, cutting baseline mean absolute error by ~40%. The improvements are projected to increase proposal sales by approximately 4%, generating an estimated $10M additional revenue.
Data Scientist at HCL Technologies
December 31, 2025 - July 18, 2025Developed an end-to-end RAG system for automated incident resolution with reproducible Makefile setup, tests, and Quarto report. Evaluated Google and OpenAI embedding models using RAGAS, achieving 78%+ faithfulness and 62%+ answer correctness. Ran LLM-as-a-judge evaluation for resolution strategies, confirming RAG outperformance via ANOVA test with significance p < 0.01. Tested embedding robustness to character-level noise with a ~4% drop in target metrics per 10% corruption.
Machine Learning Engineer at Philip Morris International
December 31, 2024 - July 18, 2025Executed an end-to-end ML project involving collection, cleaning, and preprocessing of sound data from IQOS sticks. Trained a neural network on Fourier transform features to classify original vs. fake products, achieving 86% accuracy. Submitted two reports accepted and presented to PMI research center in Switzerland.
Machine Learning Engineer at Activeloop
December 31, 2024 - July 18, 2025Modified Meta AI's MusicGen training pipeline to run using Deep Lake, collected and labeled Armenian Music Dataset, and fine-tuned the model for Armenian music generation. Authored a detailed blog about the training process and dataset preparation.
Research Assistant / Data Scientist at American University of Armenia
December 31, 2024 - July 18, 2025Analyzed the impact of detergents on chemical levels in Lake Sevan as part of an EU-funded research project. Cleaned and transformed raw environmental data, followed by exploratory data analysis and pattern discovery.
Machine Learning Engineer at Philip Morris International
December 31, 2024 - July 18, 2025Conducted end-to-end ML project for detecting counterfeit IQOS products by collecting and preprocessing sound data. Trained a neural network on Fourier Transform features to classify original versus fake products, achieving 86% accuracy. Submitted two accepted reports presented to PMI research center in Switzerland.
Machine Learning Engineer at Activeloop
December 31, 2024 - July 18, 2025Modified MusicGen by Meta AI (AudioCraft) training pipeline to use Deep Lake. Collected, labeled, preprocessed Armenian Music Dataset for training. Fine-tuned the model for Armenian Music Generation using Deep Lake Vector Store and authored a detailed blog on the process.
Research Assistant / Data Scientist at American University of Armenia
December 31, 2024 - July 18, 2025Analyzed environmental impact of detergents on chemical levels in Lake Sevan as part of EU-funded EU4Sevan program. Responsible for data cleaning, transformation, exploratory analysis, and pattern discovery in chemical data.
AI Content Creator at YouTube
January 1, 2024 - PresentFounded and manage an AI-focused educational YouTube channel, AI For Beginners, simplifying complex AI concepts with engaging visuals. Grew channel to over 18.1k subscribers and 343k+ views with a 99%+ like ratio across 30 videos, demonstrating strong audience engagement and efficient content planning. Successfully balanced channel growth with full-time studies and part-time work.
AI Content Creator at YouTube
January 1, 2024 - PresentFounder and Creator of an AI-focused educational YouTube channel (AI For Beginners) simplifying complex AI concepts for non-technical audiences with engaging visuals. Grew to 18.1k+ subscribers and 343k+ views, maintaining a 99%+ like ratio across 30 videos, demonstrating strong audience engagement. Successfully scaled the channel alongside full-time studies and part-time work.
Machine Learning Engineer at CAREAWARE
December 31, 2023 - July 18, 2025Led a small team on an end-to-end vision-based Fall Detection project using OpenCV and PyTorch. Designed an LSTM model to optimize robot arm trajectory, improving results by 50%.
Machine Learning Engineer at CAREAWARE
December 31, 2023 - July 18, 2025Led a small team on vision-based Fall Detection using OpenCV and PyTorch. Designed LSTM model to optimize robot arm trajectory with a 50% improvement. Worked onsite on projects in collaboration with Gerhard Schubert.
Education
Masters in Data Science at University of British Columbia (UBC)
January 1, 2024 - December 31, 2025BS in Data Science at American University of Armenia (AUA)
January 1, 2020 - December 31, 2024Master's at University of British Columbia (UBC)
January 1, 2024 - December 31, 2025Bachelor's at American University of Armenia (AUA)
January 1, 2020 - December 31, 2024Qualifications
Industry Experience
Software & Internet, Media & Entertainment, Professional Services, Education, Manufacturing
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Hire a Data Scientist
We have the best data scientist experts on Twine. Hire a data scientist in Vancouver today.