Tatev Aslanyan

Where Passion
Meets Will

Artificial Intelligence Solutions Architect • Experienced Machine Learning Researcher • Woman in AI • Educator • Developer

Discover my personal portfolio, a tapestry of silent achievements that embody growth and transformation. From pioneering research to transformative projects, join me in celebrating these pivotal moments. Explore and be inspired to embrace your own achievements and the endless potential of ongoing learning.

Self-Led Learning

Dive into some of my self-directed and self-motivated personal projects, Case Studies

BabyGPT: Building GPT from Scratch

Generative AI

Key Areas: Artificial Intelligence, Generative AI, Large Language Models, LLMs, Pre-Training, Transformers, Multi-Head Self Attention, Masking Mechanism, PyTorch

Predicting Stock Prices with RNNs

Deep LEarning

Key Areas: Artificial Intelligence, Sequence Based Models, Deep Learning, Time Series, Recurrent Neural Networks, LSTMs, Stock Price Prediction, Python, TensorFlow

Image Recognition with CNNs

Deep Learning

Key Areas: AI, Deep Learning, Image Recognition, Convolutional Neural Networks, Computer Vision, Python, TensorFlow

Predicting Customer Churn with ANNs

Deep Learning

Key Areas: AI, Deep Learning, Churn Rate Prediction, Artificial Neural Networks, Python, TensorFlow

Playlist Success with Machine Learning

Machine Learning

Key Areas: Product Data Science, Causal Analysis, Semantic Analysis, NLP, Linear Regression, Machine Learning, Music Playlist, Python

Fueling Success

Dive into some of the companies and organizations I collaborated with.

Where Data meets Science

Dive into few of my professional projects from my last 5 years tenure
Project 1

Next-Gen Data Integration ​

Data Solution in Cloud Integration & ETL Architecture for Cloud Cost Computing Product for Silicon Valley Start-Up

Role: Senior Data Scientist

Company: Arcurve

Client: Sillicon Valley Start-Up

Responsibilities: Overseeing Entire Data Domain, Data Identification and Acquisition, Data Engineering and Integration, ETL Architecture and Optimization, Load Testing and Data Storage, Data Processing and Analysis, Active Collaboration and Stakeholder Management, Presentations

Key Areas: Data Engineering, Cloud Computing & Integration, CI/CD, Advanced Analytics, ETL, Architecture & Design

Project 2

Harnessing Generative AI for Search Engine

Harnessing GenerativeAI for SciVal to Revolutionizing Research Analytics: Unveiling Patent Analytics through Large Language Models (LLMs) BERT Integration in SciVal

Role: Data Scientist III

Company: RELX

Client: LexisNexis, SciVal

Responsibilities: R&D, Data Evaluation and Analysis, Model Selection (NLP and LLMs), Model Testing and Evaluation, Model Selection and Implementation,  Collaboration and Stakeholder Management, Client Communication and Stakeholder Engagement

Key Areas: Generative AI, LLMs, Transformers, BERT, Text Classification, NLP, Search Engines, Research & Patent Analytics

Project 3

Enhancing Funding Recommender System

Creating Eligibility Features for Data-Driven Funding Recommender System: Amplifying Researcher Funding Experiences through Advanced Analytics 

Role: Data Scientist III

Company: RELX

Client: Funding Institutional

Responsibilities: R&D, Data Quality Analysis,  Feature Design & Architecture, Feature Testing & Selection, Implementation & Production, Monitoring & Client Feedback,  Collaboration & Stakeholder Management, Presentations

Key Areas: Advanced Analytics, Recommender Systems, Research Products

Project 4

NLP-ML infused Recommender Algorithm

Harnessing the Power of NLP and Machine Learning: Scalable, Hybrid Recommender System called LDA-LFM (Latent Dirichlet Allocation – Latent Factor Models) leveraging textual reviews, ratings and user demographics for enhanced article recommendations through advanced information retrieval


Role: Researcher

Organization: Erasmus University Rotterdam

Responsibilities: R&D, Data Collection, Model Development, Training, Evaluation and Optimization, Collaboration with Co-Author (Professor of Computer Science), Scientific Writing, Presentation

Key Areas: Information Retrieval, NLP, Topic Modelling, LDA, Matrix Factorization (LFM), Recommender Systems, Machine Learning, AI, Research

Project 5

System Failure Detection With AI

Outlier Detection in Time Series: Detecting Gas Station System Failures using  Artificial Intelligence, Deep Learning (Autoencoders with LSTMs) and Statistical Modelling

Role: Senior Data Scientist 

Company: Arcurve

Client: 7Eleven

Responsibilities:  R&D, Data Collection, Model Development, Model Training and Evaluation, Technical Documentation,  Collaboration & Stakeholder Management

Key Areas: Time Series Analysis, AI, Deep Learning, Autoencoders, LSTMs, Outlier Detection, Statistical Modelling

Project 6

ML Re-ranker for Recommender 

Harnessing Machine Learning, (XGBoost) and A/B Testing for Building a Re-ranking Model to Enhance Mendeley Suggest Recommender Algorithm


Role: Data Scientist I

Company: Elsevier

ClientMendeley Suggest, Science Direct

Responsibilities: R&D, Model Evaluation, Data Collection, Feature Engineering, Model Development, Model Testing and Evaluation, Experimentation with A/B Testing, Presentations

Key Areas: Machine Learning, A/B Testing, Recommender Systems, Research Products

Project 7

Data-driven Marketing with Machine Learning

Harnessing Machine Learning for Customer Segmentation and Retention: Crafting a Data-Driven Loyalty Card and Targeted Marketing Strategy Through Group Dynamics Prediction for the Largest Liquor Store in the Netherlands


Role: Data Science Intern

Company: NewCraft & Erasmus University

Client: Gall&Gall

Responsibilities:  R&D, Data Collection, Model Selection, Model Development, Model Training and Evaluation, Technical Documentation,  Collaboration & Stakeholder Management, Presentations

Key Areas: Machine Learning, Clustering, K-Means, RFM-Segmentations,  Dimensionality Reduction,  PCA, Decision Trees

Project 8

GeoAnalytics with Machine Learning

Harnessing Machine Learning and Advanced Analytics (Causal Analysis) for Decoding Landslide Climate Triggers and using ML (DBSCAN clustering) to detect Landslide Hotspots

Role: Senior Data Scientist 

Company: Arcurve

Client: WSP Golder Environmental Consulting

Responsibilities: R&D, Data Evaluation & Integration, Feature Engineering, Model Development, Training, Evaluation and Optimization, Collaboration & Stakeholder Management, Client Reporting, Presentations 

Key Areas: Causal Analysis, Correlation Analysis, Machine Learning, Clustering, Feature Engineering

More Case Studies

Here are Some More Case Studies I Worked on in the Past

Missing Data Imputation with Advanced Statistics

Using Single and Multiple Imputation Stats Models to Impute Missing Data

Anomaly Detection with ML (Isolation Forest)

Using Unsupervised Machine Learning To Detect Fraudulent Activity in Transactions

Robust Classification with Advanced Statistics-FastMCD

Making Linear Discriminant Analysis (LDA) Robust with FastMCD

Scroll to Top