Hi, I'm Umang Tank
A
Passionate Programmer.
About
"Talk less, Work more.."
I'm Umang Tank, a Machine Learning Engineer with hands-on experience at Samyak Infotech Pvt Ltd.
Looking for an opportunity to work in a challenging position combining my skills, which provides professional development, interesting experiences and personal growth.🤗 So, let's connect if anything about me interests you🧡
Current Focus: Artificial Intelligence for IT Operations
Education
Ahmedabad, Gujrat
Degree: Bachelors in IT Engineering
CGPA: 8.04/10
2019 - 2023
Experience
- Analyzed large and complex datasets to identify patterns, trends, and relationships to inform business decisions.
- Developed predictive models using algorithms such as logistic regression, decision trees, and neural networks to improve business outcomes.
- Utilized deep learning techniques, including neural networks, for addressing complex data science challenges in both NLP (Natural Language Processing) and CV (Computer Vision) domains.
- Implemented vector databases to efficiently store and manage high-dimensional feature vectors, significantly enhancing data retrieval and model performance.
- Conducted research and experimentation to stay up-to-date with the latest advancements in the field of data science.
- Worked on end-to-end Machine Learning project, conducting impactful Exploratory Data Analysis (EDA) and implementing feature engineering.
- Established ML pipeline, automating key processes to enhance project efficiency
- Developed a user-friendly API for seamless interaction with machine learning models.
- Implemented Object Oriented Design to improve readability and scalability Tools: Python, Machine Learning, Flask
-
Extra-Curricular Activities
- Open source contriutions:
- GirlScript Summer of Code
- 30 Days of DSA
- An active member of several Open source communities in Ahmedabad.
Projects

Developed a system to find and recognize jewelry items using advanced technology.
Accomplishments
We used YOLOv5 to identify jewelry and VGG16 for features exctraction. Then, we compared them to a images of the database, providing the best matches based on similarity. achieved 92% accuracy in recognition

MCQ Generation is an AI model which is used to generate MCQs from the given text.
Accomplishments
Fine-tuned GPT-2 on custom data to automatically generate multiple-choice questions (MCQs) from user-provided PDF and text documents

predictive model for accurate backorder forecasting and enhanced inventory management.
Accomplishments
The Predictive Model for Backorder Forecasting project aimed to develop a powerful predictive model that could forecast backorders based on historical data from inventories, supply chain, and sales. By leveraging machine learning algorithms and utilizing data from ERP systems, the model was able to classify products as either going into backorder or not

This is a NLP based ML web application which is used to classify the News Articles.
Skills
Languages




Frameworks and Tools




Database




Other




Hire me
Open for New Opportunities
Available for new data science opportunities. Experienced in machine learning, deep learning, and NLP. Let's collaborate to drive innovation and make a difference.
Contact me
Please feel to contact me on any social media Platform.