• Extracted and analyzed current keywords to build Abster's dataset for SEO, including high-volume, relevant keywords and unrelated, banned terms. Improved keyword campaign creation efficiency by 40%.• Utilized XGBoost to predict inventory needs, effectively reducing storage space requirements. Recorded and forecasted replenishment cycles to ensure continuous stock availability during high-traffic periods.• Utilized web scraping API to collect data and conducted in-depth competitor research by analyzing images, pricing strategies, market share, and keyword strategies, informing strategic decisions.• Developed a data-driven strategy using key performance metrics such as top-of-search, CTR, and ACOS. Achieved a 96% increase in sales on underperforming listings and reduced costs by 43% on underperforming campaigns.• Designed and implemented A/B testing on Amazon listings, the optimized version has a 36% increase in sales.• Developed Tableau dashboards to visualize performance analysis, brand growth trends, and user demographic patterns, delivering actionable insights and strategic recommendations to stakeholders.• Prepared prompts from existing listings to fine-tune an instructional large language model (LLM) using PyTorch, improving the creation process for listing titles, keywords, and descriptions.

• Implemented end-to-end automation of financial data ingestion and feature preparation for developing Deep Learning models such as LSTM and MLP. Improved data processing repeatability and scalability.• Maintanced ane optimized LightGBM model through feature engineering and hyperparameter tuning, validating the result and achieving a 12% increase in returns during back-test. • Successfully implemented the model in the A-share market for T0 trading, yielding a remarkable maximum annualized return of 11%

• Examined datasets from various providers in different languages and formats. Detected and resolved issues to ensure seamless integration with the company’s dataset. • Developed resampling functions based on time to enhance stock data management without compromising accuracy. Optimized the function and achieved an 80% reduction in runtime, expediting subsequent analysis.• Conducted research on accessible open-source stock market data analysis packages. Identified and integrated trending features. Translated and tailored source code for new features, aligning them with model's requirements and structure.

Prepared plasmids such as pSpCas9-2A-mCherry, CRISPR/Cas9, according to established protocols. Utilized PCR for in vitro transcription to generate gRNA. Conducted sequence verification post-preparation to ensure quality.Adhered to cell culture room’s protocols, ensuring a sterile environment. Conducted observation of cultured cells to detect contamination and assessed cytotoxicity levels regularly. Utilized Flow Cytometry for cellular uptake analysis.

Conducted inspections to verify the quality of raw materials with regulatory standards, ensuring suitability for manufacturing. Documented daily and reported anomalies on a weekly basis to enhance quality control.Employed Excel to track and manage solution ratios for various drug inspections, enhancing organization, accuracy, and convenient retriecal of data in case of investigation. Developed detail training documentation. Ensured future team members can understand the tool developed.