Fariha Tahosin Boishakhi
I am PhD student in Computer Science and Engineering at the school of School of Engineering, University of Connecticut, (Reducing Information Ecosystem Threats (RIET) Lab with a strong background in AI driven systems, data science, and computational research. My work focuses on building multimodal scalable machine learning and LLM-based solutions for real-world problems, including automated fact-checking, Safe use of AI to scientific data modeling.
Education
-
Ph.D., Computer Scinece and Engineering University of Connecticut (August 2023 - Present) -
B.Sc., Computer Scinece and Engineering BRAC University (August 2016 - December 2020)
Work Experience
PhD Summer Intern (AI) @ Regeneron Pharmaceuticals LTD (May 2025 - August 2025)
- Built a web application that automates manual verification of Data Transfer Specifications, reducing review time from 2 hours to 30 minutes.
- Engineered a document-processing pipeline to extract insights from PDFs and unstructured reference documents.
- Integrated internal APIs to generate summaries and automatically detect discrepancies.
- Designed the system for scalability and consistency, minimizing human error in document validation.
Research Assistant @ University of Connecticut (August 2023 - Present)
- Collected data of different chemical structures with API and cleaned the data to do predict eutectic temperature using Python, Beautiful Soup, Pandas, Numpy.
- Experimented on data duplication to solve imbalance data problem with random forest using scikit-learn.
- Collaborated to find new Zintl phase crystal structure based on total energy.
Business Analyst @ ACNOVA LTD. (July 2020 - Jun 2022)
- Forecasted the trend of products using previous data to import the products according to the clients need.
- Consumer market and cost analysis of cleaning and hygiene product line of FRESCO to launch new products.
- Generated leads using Google API that increase sales by 15% within 3 months
- Contributed to KPI analysis, loss elimination, and process improvement
Survey Analyst @ UNHCR – PROTTASHI (Jan 2022 - Mar 2022)
- Analysis of the Rohingya camp data in for a monthly report of cash distribution, facilities and effectiveness of the facilities given to the camp in healthcare and local cashflow
Projects
Data Duplication to solve small Imbalanced Dataset Problem with Random Forest
Explored possibility of solving small Imbalance dataset classification problem with data duplication and Random forest.
CDS&E: Collaborative Research: Designing New Zintl Phases with Motif-based Learning and Ab Initio Methods (Ongoing)
Detected Zintl phase anion and cation from chemical structure using GNN and SVM to identify structural patterns and first-principles calculations to predict and optimize novel compounds.
Eutectic Mixture Melting Point (Ongoing)
Using Siamese Network to predict different material’s mixture’s melting Point
Talks & Lectures
Publications
- F. T. Boishakhi, P. C. Shill and M. G. R. Alam, “Multi-modal Hate Speech Detection using Machine Learning,” 2021 IEEE International Conference on Big Data (Big Data) , Orlando, FL, USA, 2021, pp. 4496-4499, doi: 10.1109/BigData52589.2021.9671955.