S Dhipesh Kumar

M.Tech Graduate in Artificial Intelligence with hands-on experience building end-to-end machine learning models and AI automation pipelines using Python, APIs, and large language models (LLMs). Over the course of multiple academic and practical projects, worked on real-world use cases such as medical image classification, AI task planning applications, and workflow automation for productivity and operations. These projects include building ensemble and deep learning models, integrating third-party APIs, and deploying solutions in simple yet effective environments. Achieved measurable impact such as 98% classification accuracy in a medical imaging project and more than 40% reduction in manual work through AI-driven automation workflows.

Can help clients with the complete lifecycle of an ML or automation project, starting from data preprocessing, feature engineering, and model development to model evaluation and basic deployment. Comfortable working with evaluation metrics such as accuracy, precision, recall, and F1-score, and applying cross-validation to ensure robust performance. Additionally, experienced with explainable AI techniques like SHAP and LIME to make model behavior more transparent and interpretable for non-technical stakeholders, especially in sensitive domains like healthcare and business decision-making.

Also experienced in integrating OpenAI and Google Gemini with automation and orchestration tools like n8n to build AI-powered workflows. This includes automating repetitive tasks using Gmail API and Google Sheets API, such as categorizing and replying to emails, logging data into spreadsheets, and generating summaries or reports. Capable of designing pipelines that connect multiple services, reduce manual effort, and increase overall efficiency for small teams, freelancers, and businesses.

Looking for beginner to intermediate projects where a reliable developer is needed to contribute to practical, outcome-focused solutions. Can build classification and regression ML models, from traditional algorithms to ensemble approaches, depending on data and requirements. Able to create simple deep learning pipelines with CNN architectures such as MobileNet, DenseNet, and InceptionNet, particularly for image-related tasks. Can automate email, Google Sheets, and general data workflows using Python, REST APIs, and integration platforms. Additionally, can develop small AI tools and assistants using LangChain and Streamlit, including task planners, productivity assistants, and lightweight dashboards that make AI capabilities accessible to end users.

Services offered include ML model development (classification, regression, and ensemble models), deep learning solutions for image classification, AI automation workflows using n8n with OpenAI or Gemini, Gmail/Google Sheets/Google Drive automation through APIs, data cleaning, exploratory data analysis (EDA), and visualization using Pandas, NumPy, and Matplotlib. Also offer explainable AI (SHAP, LIME) for model interpretability and simple AI web applications using Streamlit combined with LangChain.