Prashant Verma | Data Scientist

Hi, I'm Prashant Verma.

A

Self-driven, quick starter, passionate programmer with a curious mind who enjoys solving a complex and challenging real-world problems using Data.

About

I am a Data Science professional and hold an MSc degree in Data Science from University College London with an extensive fascination for solving real-life problems using data. I enjoy problem-solving and coding. Always strive to bring 100% to the work I do.

As a Data Scientist, I possess valuable industry experience leveraging advanced data science techniques including statistics, data mining, programming, and machine learning across diverse data types. My expertise extends to web development and spans SaaS-based platforms, healthcare, and business domains. I excel in adapting to emerging technologies and expanding my knowledge base. Additionally, I have a proven track record in stakeholder management and delivering actionable business insights.

I am passionate about developing complex applications that solve real-world problems impacting millions of users.

Languages: Python, Stata, R, Java, JavaScript, C, C++, HTML/CSS, React, Bash
Databases: MySQL, PostgreSQL, MongoDB
Libraries: NumPy, Pandas, Matplotlib, Seaborn, Sklearn, Spacy, NLTK, Huggingface OpenCV
Frameworks: Flask, Django, Keras, TensorFlow, PyTorch, Bootstrap, Apache Beam
Tools & Technologies: Git, Docker, AWS, GCP, Heroku, JIRA

Looking for an opportunity to work in a challenging position combining my skills in Data Science and Software Engineering, which provides professional development, interesting experiences and personal growth.

Experience

University College London

Data Scientist

Implemented an NLP-based optimized model that extracts the important eye related and gene related information from the free-text clinical notes written by doctors at Moorfield Eye Hospital (London).
Designed the model to find the relationships between the extracted measurements and clinical details to structurely convert the information into a tabular form that separates eyeside with measurements and type of testing.
Leveraged classification and NER techniques on MEH hospital data of 5 million patients, applied various deep-learning models (RNN, LSTM, BERT).
Tools: Python, Data Science, Natural Language Processing, Flask

January 2022 - December 2022 | London, UK

University College London

Data Science PGTA

Supervise Business/Data Science students during practical sessions to perform the practical lab based on learnings from the lectures and teach new concepts.
Support module leader in designing the course content based on feedbacks and regulate assessments marking.
Assisting Modules: Python, Database Systems (Databse desinging, Normalization, mySQL), Web Technologies (HTML, CSS, JavaScript), Information Systems (Business design - DFDs, Stakeholder Analysis, MOSCOW Analysis, BPMN), Java (Object oriented programming).
Tools: Python, mySQL, HTML, CSS, JavaScript, Java, Information Systems

September 2021 - Present | London, UK

Phenopolis Ltd

Machine Learning Software Engineer

Developed web applications using react for bioinformatics and healthcare to drive and innovate medical research in machine-learning.
Carried out data research and preprocessing with machine learning and deployment team to develop the industrial projects for the UK hospitals specifically in healthcare.
Tools: React, Python, Flask, JavaScript

April 2022 - May 2022 | London, UK

University of Wolverhampton

Machine Learning Intern

Fabricated a scientific search system for Covid-19 using different deep learning approaches such as Transformers.
The model was trained on COVID-19 Open Research Dataset released by the White House.
Tools: Python, Scikit-learn, NLTK, BERT

May 2021 - July 2021 | Wolverhampton, UK

Sutra Tech Labs Pvt Ltd

Associate Data Scientist

Built automatic scrapers for populating news feed in 6 different SaaS platforms like, travel, fashion, make-up with relevant and interesting data for users. Besides, introduced hashtags to keywords in news feed, and designed a data analytical dashboard for administration.
Worked on three web applications targeting customers, selling vendors, and admin users; built RESTful APIs with functionalities such as login/sign up, view a product, add a product to cart, checkout the order, etc.
Tools: Python, NLTK, BeautifulSoup, MongoDB, Scikit-learn

June 2019 - August 2019 | Gurugram, India

Projects

                
Music Player Web-App
                  A music streaming web app based on Django
                
AccomplishmentsTools: Django, HTML, CSS, Bootstrap, SQLite
Register/login to the web app(with OAuth-based Google Sign-In).
Search and filter songs based on language and singer.
Create multiple playlists and add/remove songs to/from playlist.
Scroll through recently played/viewed songs.

VABert - ML app
                  A machine learning based VA-Bert, extracts visual acuity from free-text clinical notes of eye disease patients reported by medical specialists.
                
AccomplishmentsTools: Python, NLTK, pandas, regex, BERT, django
The model also focuses on finding relationships between extracted visual acuities aka relation extraction and later categories visual acuities as corrected and uncorrected in structured table format.
Labelled 2000 clinical letters of 1000 hospital patients by gaining domain knowledge in ophthalmology with Label-Studio tool for learning and training of models. 

Covid 19 Q/A
                  An attention-based classification model that aims at generating an answer for scientific questions related to covid.
                
AccomplishmentsIncorporated Bidirectional Encoder Representations from Transformer (BERT) that answers high-priority scientific questions with a vision to help medical community including doctors and researchers
Was trained for Q/A task from abstract of 20,000 articles related to covid-19 and got 78% overall accuracy on the
                    QA evaluation metric.

VCop - Police Station
                  A Virtual Police Station enabled with Artificial Intelligence-based smart Police officer that capture real-time reports and connects all police stations.
                
AccomplishmentsAutomatically convert complaints to legal FIR using machine learning, verification through national ID card, allowing OCR/digital signature of SHO, deploy Machine Learning (NER) model for police investigation and data analytics dashboard to track crimes in state. Incorporated CNN to detect and classify people in the video frames and Long Short Term Memory to track criminals
Accepted by government and estimated to help the state of over 72 million population.
 
VAIDYA - MEDBOT
                  Built a chatbot that can guide concerned patients by understanding and
                      assessing the patient's symptoms and identify care needed.
                
AccomplishmentsThe project was aimed to reach 4035 AYUSH hospitals all over India to support and assist patients by making hospitals digitally smart with multilingual feature.
Guide the concerned patients by understanding the symptoms that
                      the patient is experiencing and identify the care they need. If the
                      patient is found to be having some serious issue, will ask to fix an
                      appointment to the nearest hospital.

Covid Impact Analysis
                  Analysis on covid19 data to observe its impact on patients visit in the dental practices in UK
                
AccomplishmentsDental services have been majorly affected by the covid-19 outbreak since the beginning and there
seems to be no chance of recovery until June 2021.