Data ScienceNew Graduate
Data Science × New Graduate
As a new graduate aspiring to enter the data science industry, I have honed skills in building predictive models with Python and extracting/analyzing large-scale data using SQL through graduate-level machine learning research and two long-term internships at AI startups. With a proven track record of earning a Kaggle medal, my strength lies in approaching business challenges mathematically and implementing end-to-end solutions.
📌 Career Summary
- ▸Data Scientist Intern at AI Startup A (Sep 2022 - Aug 2023): Contributed to an EC site's recommendation engine improvement project, handling feature engineering and offline model evaluation.
- ▸Research Assistant at University Startup B (Apr 2023 - Mar 2024): Developed a sentiment analysis tool for customer reviews using NLP from scratch, successfully deploying it as an API.
- ▸Graduate Student, Department of Intelligent Systems, School of Engineering (Apr 2021 - Present): Conducting research on anomaly detection algorithms using time-series data, with a poster presentation at an international conference.
📌 Achievements & Strengths
- ▸Revamped recommendation model during internship, boosting click-through rate (CTR) by 14.5% and increasing monthly sales by approximately ¥8 million.
- ▸Built ensemble methods using LightGBM and neural networks in Kaggle tabular competitions, achieving top 2% (Silver Medal).
- ▸Set up and managed lab server environment (AWS EC2/GPU), automating compute resource allocation for 15 researchers, improving research efficiency by 30%.
📌 Skills
- ▸Programming Languages: Python (Pandas, NumPy, Scikit-learn), SQL, R
- ▸Machine Learning & Deep Learning: PyTorch, LightGBM, XGBoost, Hugging Face Transformers
- ▸Data Infrastructure & MLOps: AWS (S3, EC2, SageMaker), Docker, Git/GitHub, MLflow
- ▸Languages & Others: English (TOEIC 860, proficient in reading papers and writing technical docs), Statistical Society of Japan Level 2 certification
📌 Self-PR
- ▸Driving force bridging technology and business: During my internship, I faced the challenge where 'model accuracy was high but inference costs were unjustifiable.' Collaborating with engineers, I optimized the model through distillation and ONNX format conversion, reducing latency by 40% while maintaining 98% accuracy. I am dedicated to delivering end-to-end value by balancing business requirements with technical constraints.
💡 Key Tips
- 💡For new grad data science roles, emphasize both potential and implementation skills. Prepare to discuss not just Kaggle rankings, but the hypothesis testing and feature engineering behind them.
- 💡Regardless of company type, SQL and data preprocessing fundamentals are essential. Highlight your experience cleaning messy data, not just building models.
- 💡Articulate your research and internship achievements in terms of business impact (e.g., revenue, cost savings) so non-technical HR or executives can understand.
- 💡If targeting foreign companies, prepare for coding tests (e.g., LeetCode) and technical interviews in English by practicing algorithm basics and project explanations.
🔗 🔗 Other levels
Last reviewed: 2026-09-06 · aijobs Editorial