Swapnil Vishwasrao

• Lead and mentor a team of data scientists, machine learning engineers, and analysts to deliver high-impact data projects.
• Oversee team workload, project prioritization, and resource allocation to ensure timely and quality project delivery.
• Act as a technical coach, providing guidance on best practices in model development, validation, and deployment.
• Collaborate with cross-functional teams including product, engineering, and business stakeholders to align data science initiatives with organizational goals.

• Lead the development and implementation of generative AI models for various applications, including image and text generation.
• Conduct cutting-edge research in AI and machine learning to develop innovative algorithms and models.
• Collaborate with product and engineering teams to integrate generative AI solutions into products, enhancing user experience.
• Conduct thorough experiments to optimize model performance, scalability, and reliability.
• Develop technical architecture solutions for RFPs ensuring alignment with client requirements and industry best practices.
• Mentor junior researchers and provide guidance on their projects, fostering a collaborative research environment.

Projects

Field Assist Agent:

The field assist agent is a virtual assistant designed to support field engineers in their day-to-day tasks. leveraging the capabilities of agentic AI, this intelligent system provides real-time assistance, improving efficiency, reducing response time and enhance overall productivity.

AmplifiDocs

Helps enterprises mine the treasure of information from all forms of documents and enables it for downstream automation. Embedded generative AI features makes the solution generic, flexible ,contextual and effortless in terms of adoption. It is a workbench where documents can be ingested with no/minimal training. Extraction and downstream action can be orchestrated seamlessly.

SDLC Workbench

This is a cutting edge solution that enhances the traditional software development life cycle by integrating the power of AI. This innovative approach utilizes AI algorithms to automate key tasks ,improve decision- making and accelerate the software development process ,delivering high quality software with greater speed and efficiency

Custom LLM Amplifier

· This is a workbench that allows users to create a custom model using their own data.

· Users can upload their own data to train custom large language model

· This workbench is designed to help users to create models to their specific needs.

LLM As a Service

· LLM as service is an offering that allow users to access, utilize, and integrate large language models into their applications and workflows without the need to build train or maintain the models themselves.

· These services typically provide powerful APIs and tools for leveraging advanced NLP capabilities.

Car Design Ideation

Utilized my expertise to fine-tune the Stable Diffusion v2 model, an advanced and SoTA multi-modal Text-to-Image generation system, for a leading Indian car manufacturing client. This led to the development of improved car designs from textual inputs and successfully mitigated bias in the base model following the fine-tuning process.

· The following APIs have been successfully created during the development of Generative AI Studio using OpenAI Text Completion Engine (text-davinci-003) - code generation using natural language and unit test cases.

Damage Detection

· Developed custom Object Detection model that can detect and localize the damaged parts of the Car for a leading Automotive client

• Project Name: Structural Damage Detection
• Technologies: Python, DeepLearning (Tensorflow, keras), Computer Vision (OpenCV), Flask, Bit-Bucket, Nvidia Tesla V100 GPU.
• Description: Detect structural damages using deep learning and advanced computer vision techniques.
• Project Name: Asset Failure Analytics (Generator predictive Analytics)
• Technologies: Python, DeepLearning (Tensorflow, keras), Hive, Tableau, Flask, Bit-Bucket, Nvidia Tesla V100 GPU.
• Description: Asset failure predictive solution focuses on improving the overall productivity of equipment's by evaluating historic failure details, risk of failure cause of failure etc.
• Project Name: Social Media Analytics
• Technologies: Python, DeepLearning (Tensorflow, keras), Hive, Tableau, NLP, BERT, Flask, Bit-Bucket, Nvidia Tesla V100 GPU.
• Description: This project is intended to understand the customer sentiment and to find out the topic of discussion thus by analyzing these factors can understand customer needs and can improve service.

• Project Name: Decision Making Model (Process Automation)
• Technologies: Python, DeepLearning (Tensorflow, keras), USE(Universal Sentence Encoder),NLP(Spacy,NLTK,Genesim), PCF, Git
• Description: Project is designed to automate the existing ITSM process.
• Project Name: Test Automation Suite (Chatbot)
• Technologies: Python, NLP, RASA, PCF, git
• Description: This tool is designed to automate the testing process using conversational AI.
• Project Name: Entity Matching Framework
• Technologies: Python,NLP, AWS
• Description: This is the application of finding records in a data set that refer to the same entity across different data sources.

• Project Name: Wind Turbine Predictive Analytics
• Technologies: Python, Machine learning, Git, AWS
• Description: Wind energy companies need to reduce the operational and maintenance cost and downtime from turbine failure.
• Project Name: Power Generation Forecast
• Technologies: Python, Machine learning, Git, AWS
• Description: This project is mainly aim to predict the electricity unit generation by wind turbine.
• Project Name: Data Lake
• Technologies: Hadoop, Sqoop, Hive, WinSCP, CDH
• Description: Suzlon has line of business in renewable energy across the world divided country wise.

• - Understand the requirement document, technical specification document and getting clarification on the same if required.
• - Prepare test scenarios, write test cases on the given requirement and functionality.
• - Perform smoke test on each new environment/build.
• - Execute regression testing.
• - Responsible for retesting of fixed defects discover in previous build.