Sakib Hosen Himel
Werkstudent Data Science, Nesto Software GmbH
I am a Computer Science graduate currently completing a Master of Science in Data Science at Berlin University of Applied Sciences (BHT), with a rigorous foundation in machine learning, statistical modelling, time series analysis, and cloud-based deployment workflows. My academic and project portfolio spans the full data science pipeline — from large-scale data preprocessing and feature engineering through to model training, validation, and production-ready deployment on AWS and GCP. What distinguishes my profile is not merely theoretical knowledge, but a consistent record of delivering measurable, reproducible outcomes: I have built forecasting systems achieving an RMS score of 0.97, cleaned and modelled 7.7 million accident records to reach over 90% predictive accuracy, and deployed end-to-end machine learning applications on AWS SageMaker with real-time inference interfaces. Nesto's mission — building a cloud-based intelligent workforce planning solution that genuinely improves operational efficiency for its clients — is precisely the domain where I intend to contribute. The intersection of time series forecasting, algorithm development, and cloud-native data pipelines that defines Nesto's technical work maps directly onto the project experience I have built across my postgraduate studies. I bring the hands-on mentality, Python proficiency, and analytical rigour that this role demands, together with the academic depth of an active Data Science master's student. I am prepared to work independently, contribute concrete algorithmic improvements to Nesto's product, and grow as a practitioner within an agile, expert-led software team in Karlsruhe.
Experience
Talent Manager
AIESEC in Germany · Berlin, Germany
- Connected students with international internship opportunities by matching candidate profiles to relevant programmes across partner organisations.
- Supported the recruitment and selection process to ensure well-matched placements between applicants and hosting institutions.
- Organised career development events in cooperation with universities and external organisations to promote AIESEC's global programmes.
- Gained practical experience in project coordination, stakeholder communication, and cross-cultural teamwork within a non-profit, international environment.
Teaching Assistant
Daffodil International University · Dhaka, Bangladesh
- Delivered academic support to 40+ students in Artificial Intelligence coursework through structured lectures, assignment guidance, and practical sessions.
- Guided 50+ students through Database Management coursework, covering SQL query construction and database normalisation to establish sound conceptual understanding.
- Translated complex technical material into clear, accessible instruction — a discipline that informs how I document and communicate analytical work today.
Project: Hourly Power Demand Forecasting using Time Series & Machine Learning
- Irregular energy consumption readings made reliable hourly demand prediction impractical; I constructed advanced LSTM and Facebook Prophet time series pipelines with outlier removal, scaling, and K-Fold cross-validation, achieving an RMS score of 0.97 and improving model accuracy by 10% through systematic preprocessing and validation.
- Tech Stack: Pandas, NumPy, Matplotlib, Facebook Prophet, LSTM, RNN, K-Fold validation
Project: Comprehensive Analysis and Prediction of US Traffic Accidents
- Quality issues across 7.7 million U.S. accident records (2016–2023) undermined the reliability of severity prediction models; I led end-to-end data preprocessing to improve data quality by 25% and applied AutoML-based machine learning models to predict accident severity with over 90% accuracy, surfacing actionable insights into accident risk factors.
- Tech Stack: Python (Pandas, Matplotlib, Seaborn, Plotly, GeoPandas), AutoML
Project: Flight Fare Forecasting — Exploratory Analysis & ML Deployment
Personal Portfolio Project
- Volatile flight pricing data required a reproducible, production-grade ML pipeline with real-time inference capability; I designed custom scikit-learn feature engineering pipelines, deployed the trained models on AWS SageMaker and EC2, and built a Streamlit application for live fare prediction, with AWS S3 handling data storage throughout.
- Tech Stack: Python (NumPy, Pandas, Matplotlib, Scikit-learn), AWS SageMaker, EC2, S3, Streamlit
Research: Perception of Multimodal Objects in NLP through Computer Vision
- Designed and implemented a real-time, voice-interactive AI application that integrated speech recognition, computer vision, and natural language processing, enabling users to identify and classify physical objects via camera input with audio feedback delivered through text-to-speech synthesis.
Education
Master of Science
Data Science · Berlin University of Applied Sciences (BHT)
- Relevant coursework: Machine Learning, Deep Learning, Neural Networks, Natural Language Processing, Computer Vision, Statistical Computing, Big Data Architecture, Cloud Computing, Data Visualisation, and Business Intelligence.
- Practical experience with Python, R, NumPy, Pandas, Docker, and NoSQL databases through applied data science workflows including data preparation, cleaning, and integration.
- Portfolio of independently executed ML and data engineering projects spanning time series forecasting, image classification, NLP, and cloud-based deployment.
Bachelor of Science
Computer Science and Engineering · Daffodil International University (DIU)
- Relevant coursework: Data Structures & Algorithms, Artificial Intelligence, Data Mining, Machine Learning, Database Management Systems, Digital Image Processing, Computer Networks, Operating Systems, Compiler Design, Object-Oriented Programming, Software Engineering, and Computer Architecture.
- Practical programming experience with C, Java, and Python through applied projects and programming competitions.
Skills
Python (NumPy, Pandas, Scikit-learn, Matplotlib) · Machine Learning & Predictive Modelling · Time Series Analysis (LSTM, Prophet, RNN) · Data Preprocessing & Feature Engineering · SQL & Database Management · Cloud Deployment (AWS SageMaker, EC2, S3, GCP) · Deep Learning (TensorFlow, PyTorch, Keras) · Data Visualisation (Plotly, Streamlit, Seaborn) · Statistical Analysis & Model Validation · Docker & Data Pipeline Architecture