Mihir Joshi

Data Scientist at The Home Depot

Atlanta, Georgia, United States

About

Motivated analyst with a demonstrated history of working on Data Science projects. Skilled in Python, SQL, Tableau and cloud services like AWS, GCP and MS Azure. Experienced in using various SQL and No SQL databases. Strong consulting experience with a Masters focused in Information Management from School of Information Sciences, University of Illinois at Urbana-Champaign.

Experience

  • The Home Depot ()
    • Data Scientist
      Oct 2024 - Present · 1 yr 10 mos

    • Senior Data Analyst
      Oct 2022 - Oct 2024 · 2 yrs 1 mo

    • Data Analyst
      Dec 2021 - Oct 2022 · 11 mos

  • Data Analyst at Capital One
    Mar 2021 - Dec 2021 · 10 mos

  • Machine Learning Analyst at University of Illinois at Urbana-Champaign
    Jul 2020 - Mar 2021 · 9 mos

    • Determined the best alternative for a natural language processing solution by performing a comparative analysis on various cloud services like Azure Machine Learning, Amazon Comprehend and Google Cloud Natural Language • Analyzed large library datasets by researching topic modeling algorithms like Latent Dirichlet Allocation • Employed the best topic modeling alternative to assist researchers expedite scoping and systematic reviews

  • Data Science Intern at Corteva Agriscience
    Sep 2019 - May 2020 · 9 mos

    • Evaluated the effects of the features from MODIS data on Google Earth Engine on crop yield prediction • Predicted crop yield for corn by building machine learning models using XGBoost and Random Forest regressors • Collaborated on the GCP AI Platform and discovered that a fine-tuned XGBoost improved previous accuracy by 16% • Aggregated and integrated features of SMAP soil moisture data from Google Earth Engine to the existing dataset • Identified the impact of SMAP features by performing feature importance analysis and visualizing the results in QGIS

  • University of Illinois at Urbana-Champaign ()
    • Data Engineer
      Dec 2019 - Apr 2020 · 5 mos

      • Streamlined the process of analyzing actions and interactions of students with 3D objects in Virtual Reality classrooms • Laid foundation to the analytics stage of the project by building a machine learning pipeline in Azure Machine Learning • Defined metrics like object interactivity and student mobility during a lecture that provides focus points to lecturers • Stored all interaction data along with the student engagement metrics in PostgreSQL to achieve fast query responses

    • Graduate Research Developer
      May 2019 - Dec 2019 · 8 mos

      • Developed an open-source web application to catalog large-scale visualizations like 3D models and 360° videos • Hosted the website on Docker with a Python Flask backend and built an interactive framework using Vue.js • Stored service and endpoint related data on MS SQL Server, and metadata of all visualizations on MongoDB • Published various visualizations on suitable platforms like Sketchfab for 3D models and Vimeo for 360° videos