I have been involved to come up with ML solutions for the company and automate manual processes of existing software. I also have to present valuable insights from the data to the technical and non-technical teams.The first project I worked on was ITScope Catalogue Automation, where all the products registered by our clients were automatically assigned using ML model to one of the 142 categories provided. The final model was built using XGBClassifier which achieved an accuracy of 92% with an 82.3% of F-Score. I developed an end-to-end ML pipeline that involves technologies like Python, Pandas, Numpy, NLP, Nltk, Sklearn, XGBoost, Transformers, Gensim, Optuna, FastAPI, Docker, Git, CI/CD, PostgreSQL, Tableau, Seaborn, Matplotlib and Jupyter Notebook.The second project I'm working on is Item Return Shipment, where the model has to provide the predictions of the items/products ordered by the customer, which could be returned.

Created multi-class classification of the comments at the NDR News portal. A total of 9 classes are categorized based on the sentiment of commenters and offtopic to the news published. The final model based on Neural Network has achieved 93% of accuracy with 82% of F-Score. The second project I worked on was Multimodal Application for Second Language Learners, where the end product will help the users by showing the relevant images for the complex words identified. I have been involved in collecting data from the web and creating a complex word identification model for English language learners. For this, I have created 2 final models based on Bi-LSTM and Transformer based models by fine-tuning BERT and RoBerta. I have also incorporated many external features that could help with complex word identification. The performance achieved by the model was 95% of accuracy with an 87% of F-Score, which outperforms the previous state-of-the-art results achieved on those Datasets.

Other than requirement gathering, analyzing functional specification, development, and maintenance of the Hospitality management system, I have created the machine learning model for Customer Market Segmentation using Hierarchical Agglomerative clustering algorithms and achieved a 0.57 Silhouette Score.

I have worked in 3 different domains including ERP systems, Claim Management Systems, and Transport Management Systems. I have managed client interactions and conducted comprehensive requirement analysis, ensuring alignment between project deliverables and client expectations. I have been involved in requirement analysis, software development, unit testing, and providing support. Achieved experience in development using Agile methodology. I have also mentored junior developers and provided onboarding training for new team members, fostering a collaborative and knowledge-sharing environment within the project team.

Developed an E-Commerce web application prototype for client utilizing Java and Oracle Database.