Akshita M

Data Scientist with 1.8 years of experience building data-driven solutions across manufacturing, e-commerce, and supply chain domains. Expertise in working with structured and unstructured data, merging, cleaning, building machine learning models, and delivering dashboards standardizing datasets to derive actionable insights. Adept at working with clients and stakeholders to present analytical findings, articulate the limitations of data, and collaborate across teams to resolve data quality issues and improve data practices.

Data Scientist at MATHCO (2024-8 – Present)

Gas Turbine Coating Defect Analysis (POC)

• Studied gas turbine manufacturing processes and built a data framework to identify why Thermal Barrier Coatings (TBC) develop cracks — analyzing 15+ process variables to find the ones that cause the most defects.
• Built and compared multiple machine learning models (including Logistic Regression and XGBoost) to predict coating failures before they occur, helping the team focus attention on the highest-risk process stages.
• Used clustering techniques to group defect patterns across turbine cells, surfacing hidden trends that were not visible in standard reporting.
• Designed a Power BI dashboard that gave engineers a live view of defect patterns by turbine cell, reducing reliance on manual inspection reports and supporting efforts to cut material waste.

Data Analyst at MATHCO (2025-05 – 2025-09)

Tariff Impact & Conversation Analytics

• Wrote scripts to scan and analyze 1M+ buyer-seller messages for keywords related to tariff concerns, surfacing patterns in order cancellations, pricing complaints, and product substitutions.
• Built a Power BI dashboard that let leadership filter conversation trends by market, region and keyword, making it easy to see which geographies were most affected by trade-related disruptions.
• Tracked how tariff-related complaints mapped to different stages of the order process, helping the business understand exactly where friction was leading to refunds or lost customers.
• Produced recurring trend reports summarizing how sellers were adapting their strategies, giving leadership a consistent pulse on market health during a period of trade uncertainty.

Data Analyst at MATHCO (2024-12 – 2025-04)

Inventory Optimization

• Built a recommendation system that automatically calculated ideal stock levels for battery parts across regional dealers, using historical sales data and live inventory feeds as inputs.
• Grouped dealers by how quickly they sell through stock so that high-demand regions always had enough parts while low-volume areas weren't sitting on excess inventory.
• Created a simple decision output (Add / Keep / Remove) for each dealer's stock items, turning complex data analysis into clear, actionable guidance for regional supply chain teams.

Bachelor of Engineering in Electronics and Communication – BMS Institute of Technology and Management (2019-2023)