• Hi!
    I'm Abhiram

    A Data Scientist with experience in Real Estate and Healthcare Industries. Learning everyday so my models can learn better.

    Through this portfolio I would like to give the recruiters and hiring managers a better understanding of who I am and what kind of work interests me.

About Me

Who Am I?

I'm Abhiram Satpute, an avid Data Science learner with experience as a Data Scientist at hBits Online Platform Services in Mumbai and mBreath Technologies at IIT Kharagpur. With my experience in data science, fast and dynamic start-up culture, team management and mentorship, assistance in business development, customer profiling, communication and networking, I believe I have the knowledge, drive, and technical skills needed to be the perfect addition to any data science team.

I am an honest, hard working, critical thinking, risk-taking, disciplined team worker who is always keen to learn new skills while exhibiting great time management skills. I have worked as a Data Scientist at startups comprising of elite individuals, hBits (team of around 12, with grads from Harvard, Stanford, Duke, IIT Madras and Bombay, IIM Ahmedabad and Indore, NIT Nagpur, co-founder of Shaadi.com, and other entrepreneurs and individuals), and mBreath (team of 4 with myself being the first full-time hire, and two IIT Kharagpur PhD Scholars and an IIT Kharagpur Professor, along with interns from IIT Kharagpur, NIT Raipur, and VIT).

I am happy to let you know
that 20+ ML projects are done successfully!

Find here
What I do?

My proficiency and Expertise

Python and Machine Learning

Experience of 4+ years in Python and 3+ years in Data Science. Worked on ML, PCA, DL, NLP, Scraping, Translation.

Deep Learning and NLP

I have used CNN, LSTM, BLSTM, CNN-LSTM, NLTK, spaCy, Amazon Textract, coreNLP in my projects.

Exploratory Data Visualization

EDA makes me understand the data before applying any algorithms. I have used this in almost all projects.

Enhanced Feature Extraction

I have extracted features in Temporal, Spectral, Energy, Harmonic, and Cepstral domains for good accuracy.

Intelligent Feature Selection

All features are not important, sometimes few need to be removed. It increases training speed and accuracy.

Android Apps with ML Model

Generally, I use my trained model in real-world use cases with TensorFlow Lite, Protobuf files, and Android Apps.

IoT, GIT, AWS, and GCP

I manage codes with a large team on GIT. Large ML models are trained on AWS and GCP, maintained on GitLab, and can be stored on S3.

Signal Processing and Analysis

I understand the basic math behind the signals which helps me to decide the processing required for signals such as Sound and Medical data.

Game Development

In early days, I developed a performance-based Car Racing Game (video) in Python, and a First-Person Shooting Game (video) in C# and Unity3D.

My Specialty

My Skills

I am equipped with diverse skills by the virtue of my hard works. In the past, I have worked on multiple additional roles, including Business Development, Market Research, Customer Segmentation, and Satisfaction analysis apart from my Data Science projects. These additional skill sets helped me to be a team player and builder. Besides, I have a specialty in Machine Learning and related domains.

Few of my skills are

Python

90%

Machine Learning

95%

Deep Learning

95%

Natural Language Processing

85%

Feature Selection

80%

Feature Engineering and Extraction

95%

Signal Processing with Python

75%

Android Application

70%

Cloud Processing

80%

Game Development

75%

Market and Business Research

70%

Internet of Things

60%
Experience

Work Experience

Data Scientist, hBits Online Platform Services Sept 2019 - Jan 2020

  • Mapped the Real Estate data of a whole state (which can be scaled to cover even larger areas) of India. Collected loads of structured and unstructured data using scraping algorithms and OCR, worked on their accurate translations, and applied NLP techniques on the final files using SpaCy, NLTK, and coreNLP to extract the required insights from these web pages and texts.
  • Also applied AWS Services such as (Amazon Textract, S3, and more) and NLP algorithms to convert other unstructured datasets in the form of images and PDF Documents into structured spreadsheets and tables to get better insights regarding pricing, amenities, location, and other essential factors from the scraped Real Estate data.

Data Scientist, mBreath Technologies Sept 2018 - July 2019

  • Researched and Developed three data science projects using Deep Learning (CNN, LSTM, CNN-LSTM, Inception, VGGs), Machine Learning models (Logistic Regression, Random Forest, SVM, and XGBoost), and Feature Selection for real-time audio analysis and classification, sleep apnea detection, and sleep stage classification.
  • Worked on more than 200GB of Sound data and 500GB of Medical Data on Google Cloud Platform (GCP) using multiprocessing, CUDA, Tensorflow-GPU, and Onedrive.
  • Worked on assisting the CxOs at multiple fronts including Business Model Development, Customer Acquisition, Market Analysis, Product Feasibility Analysis and Customer Segmentation and Satisfaction analysis while working as Data Scientist.
  • Exporting these projects and deploying them in Android Applications. Doing the fun-sized Java and Android programming as required.
  • Drafting these projects as research papers as the co-author using LaTeX while at IIT Kharagpur.

Application Development Associate, Accenture May 2018 - Sept 2018

  • Underwent training on Java, SQL, Javascript, and HTML/CSS. Passed with flying colors.
  • Excelled in the final Assessment project in advanced Java. Got immediately assigned to a project as an AngularJS developer.
Competitons and Programmes

Training

  • Titan Technology Tune-in 2.0, Tamil Nadu

    April 2019 – May 2019

    This forum enabled us to showcase our product and technology to TITAN senior management and executives. The forum helped us to discuss and demonstrate our project and research work and talk about applications, growth, and use cases that are to be considered while entering the market. The meet was full of advanced researches and projects led mainly by PhD scholars and professors representing elite institutes across India (IIT Kharagpur, Kanpur, Guwahati, IIIT Jabalpur, etc.).

  • Health Trail Programme, IIM Calcutta

    Jan 2019 – April 2019

    Health trail was a three month non-residential accelerator programme for healthcare startups hosted by IIM Calcutta and Villgro. Got to learn about brand development, market research, GTM strategies, balance sheet handling, agreements, and all round business knowledge required to grow a startup while networking with entrepreneurs all around the country.

  • Qualcomm Design In India Programme

    September 2018 – March 2019

    This was one of the toughest startup programme to get selected. After the selection in the final round, I learned a bit of Android Kernel programming, sensors interfacing with Qualcomm's Snapdragon processors, Java Native Interface (C and Java coding together), while researching and developing the AI models simultaneously. Final event of Top 15 Nation-wide product-based startups was venued at The Ritz Carlton, Bangalore. No picture to display (other than my profile picture on the left) due to immense work and busy timeline.

Academia

Education

During my journey as an Engineer at IET DAVV, I developed basic projects in Java, Android Apps, and Games from scratch. Later, I underwent training in Hadoop and HDFS, and finally started working on Data Science projects by the third year. Cleared my Engineering with First Level with Distinction.

Apart from technical and academics, I also co-founded an art club in college called The Art Legion. We organized as well as participated in events that raise social awareness through art.

I also captained the branch football team for 3 years. Represented college team in IIM Indore (Runners-up), BITS Goa (twice), and DAVV Inter-Divison tournament (Winners).

Passed 12th with 87.4% and 10th with 9.0 CGPA. Represented the school in school and district-level football tournaments. House Vice-Captain.

Get in touch

Contact

I would like to thank you for visiting my page and your attention. Please visit my LinkedIn web page to contact me.

Other Profiles:

Stack Overflow   Youtube   GitHub   Twitter   Hackerrank