Pratik K.

About Me

Results-driven Data Scientist and GenAI Engineer with over 7 years of experience specializing in Machine Learning, Deep Learning, NLP, Computer Vision, and LLM-based applications. Demonstrated success in delivering end-to-end solutions on cloud platforms, resulting in significant business impact such as $14M savings through predictive maintenance and 80% efficiency improvements via knowledge graph implementation. Proficient in Python, TensorFlow, PyTorch, Azure, GCP, and OpenAI technologies.

AI, ML & LLM

Generative AI Deep Learning LangChain Machine Learning OpenAI Pytorch Vertex AI

Backend

Database

DevOps

CI/CD AWS Azure GCP

Workflow

Other

Work history

Cognizant
Cognizant
Senior Associate - Projects
2024 - Present (1 year)
Mumbai, India
  • Spearheaded a comprehensive knowledge graph integration using FalkorDB, incorporating Python and OpenAI APIs within the client's proprietary agentic framework using Agile methodology in just 4 weeks, resulting in an 80% surge in developer efficiency.

  • Delivered a feature to generate architecture overviews and README files using the generated knowledge graph within repositories, completed within 2 weeks.

PythonOpenAI APIs FalkorDB Knowledge Graphs DockerLangChain Langgraph CrewAI
Tata Consultancy Services
Tata Consultancy Services
Data Scientist
2022 - 2024 (2 years)
Mumbai, India
  • Conceptualized and executed a predictive maintenance application on Azure Web Service for asset battery degradation using Python, Dash, Plotly, and Flask, resulting in $14 million cost savings in optimized battery maintenance.

  • Guided 129 participants in TCS Youth Employment Program as an SME, helping them launch careers in Data Science.

  • Developed a predictive maintenance solution to enhance asset utilization using Fbprophet, NLP, and statistical analysis in Azure ML Studio.

  • Built a Python web app in Azure, integrating Kusto DB with Pandas, Flask, and Dash, to prioritize work orders.

  • Employed the DBSCAN algorithm in scikit-learn for predictive maintenance in the Oil & Gas sector using Azure Databricks and PySpark for daily forecasts and integrating APIs for automated predictions across 144 wells.

PythonDash PlotlyFlaskData ScienceStatistical Analysis NLPAzure ML Studio Kusto Query PandasScikit LearnAzure DatabricksPysparkAPIsDockerQuantitative Analysis AnalyticsMentorship & Coaching
Searce
Searce
Machine Learning Engineer
2021 - 2022 (1 year)
Pune, India
  • Developed a background removal application for high-resolution images (864x864 pixels) using PyTorch and U-Net architecture, resulting in a $30,000 project acquisition from the client.

  • Incorporated a unique hybrid loss function for U2-Net based on research paper analysis, resulting in heightened image processing efficiency and enhanced accuracy in image segmentation.

  • Enhanced predictive performance by engineering Vertex AI pipelines in GCP Cloud for automated training and deployment, achieving an R2 score of 0.79 on a 15-year test dataset, using Python and Google AI platform SDKs.

LTI Infotech (now LTIMindtree)
LTI Infotech (now LTIMindtree)
Consultant | Software Engineer | Trainee
2018 - 2021 (3 years)
Mumbai, India
  • Managed security requests for the client within the JDE ERP system, ensuring compliance by meticulously reviewing evidence and documents.

  • Successfully resolved over 500 support activities, maintaining a high standard of security and compliance.

  • Ensured audit readiness i n processing security requests.

  • Developed a pivot table functionality in Angular.

JDE ERP system Microsoft ExcelPivot Tables Angular

Education

Career Essentials in Generative AI
Career Essentials in Generative AI
Microsoft
2024 - 2024
MSc Data Science
MSc Data Science
Liverpool John Moores University
2022 - 2023 (1 year)
PG Diploma, Data Science
PG Diploma, Data Science
IIIT Bangalore - India
2020 - 2021 (1 year)
BE Computer Engineering
BE Computer Engineering
Vidyalankar Institute of Technology - India
2014 - 2018 (4 years)