Yogendra Verma

Hi, I'm Yogendra Verma.

A

Resourceful Data Scientist with expertise in Python, machine learning, deep learning, and NLP, skilled at unraveling complex challenges with analytical finesse and innovation

About

As a Data Scientist, I thrive on problem-solving and coding, leveraging a diverse skill set spanning Python, Machine Learning, Deep Learning, NLP, Computer Vision, Image Processing, and Cloud technologies such as AWS and GCP. My passion lies in developing AI/ML applications and conducting Quantitative & Qualitative Analytics to derive valuable insights, addressing real-world challenges with data-driven solutions.

Languages: Python,Embedded C,
Databases: MySQL
Libraries: NumPy, Pandas, OpenCV,, Keras, Tensorflow , Sklearn, Numpy, Matplotlib, RegEx, NLTK, Gensim, SpaCy, Plotly
Frameworks: Flask, Keras, TensorFlow,PyTorch
Tools & Technologies: Git, Docker, AWS, K8s Cluster, GCP, Heroku, MATLAB,

Always seeking to work with reputed organization where I could get an opportunity to grow my skills and I can upskill the organization’s productivity using my information & Knowledge.

Experience

Gramener

Sr. Data Scientist

Design and develop an AIML-based Auto Adjudication System to automate the medical claim process to make the system fast and efficient and also reduced the manual work.
Designing and deploying APIs with CI/CD pipelines and Docker for scalability and Docker building for efficient containerization and application deployment.
AWS deployments, managing applications, and optimizing solutions using K8s clusters to enhance performance, capable of handling 100,000 documents per hour
Tools: AIML, Deep learning, Computer Vision, Tesseract, Python, API Design, Flask, Dcoker, K8 cluster, AWS, Documents Classifcation, OpenAI

March 2023 - Present | Hyderabad, India

Thinkbridge

Sr. Data Scientist

Design and develop an e-billing assistant machine learning model which can automatically identify or flag the bills that have the potential to be rejected.
Design and develop an automation bill remediation system for the rejected lines Items In the e-bills to quantify the current rejection rate and help to reduce the rejection ratio.
Building an AIMl based model to identifying the fraud while processing the bills by the leagal firms to its customers. Also it is is removing maximin human effort.
Tools: Python, OpenCV, Keras, Tensorflow, PyTorch, BERT, Word Embeddings, Text Classifcation, AWS textract, Pandas Profiling, SpaCy, NLTK, Gensim,SHAP,Lime, Seaborn

June 2022 - Feb 2023 | Pune, India

Quatrro

NLP Platform Engineer

Designing and constructing a document classification system to automate the loan underwriting process.
Clustering documents from comprehensive loan packages, comprising over 1500 pages and spanning 30+ categories concurrently.
Extracting data from scanned PDF bank documents, including tables and images, to expedite the loan application underwriting process.
Tools: Python, Layout Detection, Documents Classification, BERT, Random forest, CNN, RNN, RCNN, LSTM,Pandas, NumPy, Image Processing, OpenCV, Computer Vision, Camelot,AWS textract,NER,paCy, NLTK, Gensim,SHAP,Lime, Seaborn, Univariate and multi-variate Analysis

Dec 2020 - Dec 2021 | Gurugram, India

Edu-fabrica

Data Scientist

Tools: Machine Learning, Python, Pandas, Sklearn, Natural language processing, NLTK, RegEx, Stemming, Lemmatization, CountVectorizer, TF-IDF, WordCloud, Selenium web scraping, Classification Models, Sentiment Analysis, Topic Modeling, Text Summarization, Selenium, Huggingface
Design and building an Automation system based on AIML to Support Process and costomer Reviews Sentiment Analysis and Reviews Summarization, YOLO

Aug 2018 - Nov 2020 | Delhi, India

Kyrion

Embedded System Engineer

Tools: IOT, Raspberry Pi, NodMCU AWS, MQTT, Node Red, Pub Sub, http, html,php, webhook
Design and develop IOT Based Smart home Automation To control appliances at fingertips from anywhere in the world
Additionally, developing both major and minor academic Embedded Systems and IoT projects

May 2014 - Jan 2016 | Delhi, India

Robosapiens

Research Engineer

Tools: Embedded System , Embedded C, Raspberry Pi, USART, MQTT ,Microcontrollers, Microprocessors, Proteus Tool, Raspberry Pi OS
Design and engineer Embedded Systems for specialized tasks and Smart Automation applications.

May 2012 - July 2013 | Delhi, India

Projects

                
Auto Adjudication System
                  An AIML based model to classify medical documents
                
                Accomplishments
                Tools: AIML, Deep learning, Computer Vision, Tesseract, Python, API Design, Flask, Dcoker,
                    K8
                    cluster, AWS, Documents Classifcation
AIML-based Auto Adjudication System to automate the medical claim process to make
                    the system fast and efficient and also reduced the manual work.
Designing and deploying APIs with CI/CD pipelines and Docker for scalability
                    and Docker building for efficient containerization and application deployment.
AWS deployments, managing applications, and optimizing solition using K8s cluster to skill up the
                    performance.
Tested the model on the real documents and got 95% overall accuracy on the evaluation
                    metric.

Document Quality Index Calculator 
                  To calculate Document Quality Index using Computer Vision
                
                Accomplishments
                Tools: Python, OpenCV, PyTorch, Computer Vision, Image Processing
Calculate Document Quality Index using Computer Vision
Transforming a document quality solution into an API.

E-billing Assistant Automation System 
                  An AIML based E-billing Assistant Automation System
                
                Accomplishments
                Tools: Python, OpenCV, Keras, Tensorflow, PyTorch, BERT, Word Embeddings, Text
                    Classifcation, AWS
                    textract
An AIML model to flag potential bill rejections automatically.
A system to quantify rejected items in e-bills and reduce rejection rates.
An AIML model to detect fraud in legal firms' bill processing, minimizing human effort.

E-bill remediation system
                  E-bill remediation system to quantify rejection rates.
                
                Accomplishments
                Tools: Python, OpenCV, Keras, Tensorflow, PyTorch, BERT, Word Embeddings, Multi-label Text
                    Classifcation, AWS textract
An automated bill remediation system to quantify rejection rates and reduce ratios.
A solution to assess rejected items in e-bills, aiming to give reasion of rejection 
An AI-powered fraud detection system for legal bill processing, streamlining operations.

Classification Solution for Scanned Documents
                  Document Classification Solution for Scanned pdf files for Loan application
                
                Accomplishments
                Tools: Python, Documents Classification, BERT, Random forest, CNN, RNN,
                    RCNN,
                    LSTM,Pandas, NumPy, Image Processing, OpenCV,AWS textract
Designing and constructing a document classification system to automate the loan underwriting
                    process.
Clustering documents from comprehensive loan packages, comprising over 1500 pages and spanning 30+
                    categories concurrently.
Tested the model on the real scanned documents and got 90% overall accuracy on the evaluation
                    metric.

Scanned Documents Data Extraction
                  A Documents Data Extractor to get document's data
                
                Accomplishments
                Tools: Python, Layout Detection CNN, RNN,RCNN,
                    LSTM,Pandas, NumPy, Image Processing, OpenCV, Computer Vision, Camelot,AWS textract
To extract data from PDF scanned bank documents using Deep Learning
Extracting data from scanned PDF bank documents, including tables and images, to expedite the loan
                    application underwriting process.