Welcome to my Portfolio!
Jyoti Prakash Maheswari
About Me

About Me


I'm a graduating masters student in Data Science at the University of San Francisco in June 2019. I am a seasoned problem solver and have 3+ years of experience of solving complex business problems using Data Science and Machine Learning. I have worked across varied domains and in disparate roles ranging from Operations, Quality Management, Data Analysis, and Product Management. I am interested in solving critical business problems using Machine Learning and Data Science.

Projects

Leanplum In-App Purchase Prediction

Using Customer's historical data, make prediction on probability of purchase in next 7 and 14 days.
[Python, Sklearn, Boosting, Random forest, Feature Engineering , Ensembling]

WSDM - KKBox's Music Recommendation Challenge

Predicting if the user will re-listen to a given song
[Python, Sklearn, Model Benchmarking, Recommendation, Ensembling]

LANL Earthquake Prediction

Time to next Earth-quake based on acoustic data
[Spark-ML, Spark-SQL, Python (pandas, numpy, scikit-learn), MongoDB, S3, EMR]

Fine-grained image classification

Fine-grained image classification on Caltech-UCSD Birds 200 dataset using Bi-Linear CNN and ResNet34
[Pytorch, Opencv, matplotlib, ResNet34, BCNN]

Isolation Forest for Anomaly detection

From scratch implementation of Isolation forest algorithm for Anomaly detection and incorporated changes to make it robust to noisy features.
[Python, Trees, matplotlib]

Article Recommendation

Article recommendation engine based on article similarity built using Flask.
[Recommendation Engine, Word2vec, TF-IDF, GloVe, Flask, AwS]

Experience



  • Dec 2018 to Present
    Data Science Intern, Trulia

    Working on a wide range of Computer Vision problems ranging from multi-label classification, localization and object-detection using weakly label data for Real-estate images. Benchmarked various novel model architectures and loss functions for weakly label data.

  • July 2018 to June 2019
    M.S. in Data Science, USF

    I am pursuing Masters' degree in Data Science at the University of San Francisco. Coursework includes Machine Learning, Statistical Modeling, Data Acquisition, Distributed Computing, Time Series Analysis, Experimental Design, Deep Learning, Relational & NoSQL Databases.

  • June 2017 to June 2018
    Sr Data Analyst and Product Manager, Squad

    Used data science and machine learning techniques to solve critical business problems relating to contractor management, quality, forecasting and automation. Also, worked as Product Manager for Machine Learning projects and solved mission critical problems

  • June 2016 to Jan 2017
    Assistant Quality Manager, ITC

    Lead a team of 35 employees to drive innovation, product consistency and quality through a workforce of over 800 unionized employees. Use structured problem solving approach, six sigma techniques and statistical tools to ensure superior product quality.

  • July 2015 to May 2016
    Assistant Project Engineer, ITC

    Worked on developing future ready products and ensured pilot scale run-ability and scale up. Also evaluated various technologies and processes to ensure competitive advantage and superior product quality.

  • May 2014 to July 2014
    Summer Internship, University of Pittsburgh

    Studied dehydration mechanisms, co-related Carbenium Ion Stability (CIS) with Proton Affinity (PA) of alcohols in a unique attempt to correlate CIS vs PA and study its effect

  • May 2013 to Jun 2013
    Summer Internship, IIT Guwahati

    Worked on binary linear programming technique. Implemented the algorithm relating to Timetable scheduling to derive optimum solutions using CPLEX.

  • July 2011 to May 2015
    B.Tech Chemical Engineering, NIT Warangal
    Join Secretary of ChEA, Event Manager @Chemstorm

    Undertook courses related to Mass Transfer, Heat Transfer, and fluid mechanics etc. Also took elective courses in problem solving and object oriented programming.

BLOGS and TUTORIALS

Breaking the curse of small datasets in Machine Learning: Part 1

Why the size of the data matters and how to work with small data?

10 not so intuitive things about programming with R

How to effectively use R for Data Science

Implementing Machine Learning Algorithms from Scratch

Understanding how various ML-Algorithms work under the hood