Siddhant Kadam

Data Scientist with 3 years of experience specializing in AI/ML solutions using AWS and Azure services. Proven expertise in developing and deploying Generative AI (Gen AI) solutions, fine-tuning large language models (LLMs), and creating scalable machine learning models. Proficient in AWS SageMaker, Azure Machine Learning, and managing big data with AWS S3 and Azure Blob Storage. Skilled in MLOps, implementing CI/CD pipelines using AWS CodePipeline and Azure DevOps. Strong background in serverless computing with AWS Lambda and Azure Functions. Experienced in designing end-to-end AI/ML pipelines, ensuring robust, secure, and efficient AI-driven applications.

Project: Marketing Content Validator
Role: Data Scientist and Product Owner
Description:

• Developed a Generative AI (Gen AI) solution to evaluate marketing content for auditing personnel in Asset Management Companies (AMCs).
• Led a team of four to create a product that validates content, ensuring desired rules and disclaimers are incorporated in images, PDFs, and videos.
• Integrated both generic and custom rules to perform content validation through a user-friendly interface.
• Ensured cloud flexibility by utilizing AWS Bedrock, Azure Open AI, and Google Gemini services.
• Deployed the solution using Docker, enabling efficient and scalable hosting.

Project: Email Attachment Parsing Automation Pipeline for Insurance Brokers

Role: Data Scientist

• Description:Led a five-member team to develop an automated pipeline for validating email attachments in insurance policy requests.
  Spearheaded the integration of innovative technologies, focusing on AWS Lambda for automating email processing and reducing the need for dedicated servers.
  Ensured seamless client communication to align the project with expectations and address challenges.
  Integrated Generative AI (Gen AI) models to effectively parse diverse data from email attachments, enhancing accuracy and efficiency.
  Continuously refined testing methodologies to improve robustness and reliability.
  Significantly streamlined the policy request workflow, reducing manual intervention for insurance brokers.

Project: Hireeaze - Recruiting Tool
Role: Data Scientist (Lead)
Description:

• Led a team to develop an AI-driven tool using Large Language Models (LLMs) to automate and enhance hiring workflows.

Key Achievements:

• AI-Powered Screening: Introduced LLMs for autonomous candidate screening, reducing manual effort and improving matching accuracy.
• Candidate Ranking Algorithm: Developed an algorithm to efficiently rank candidates, streamlining the shortlisting process.
• Interactive Dashboard: Designed a real-time dashboard for hiring managers, offering an intuitive user experience.
• Feedback Loop: Implemented a mechanism to continuously improve AI predictive capabilities based on user insights.
• Customization and Scalability: Ensured tool adaptability for various departmental needs, enhancing versatility.
• Data Security: Maintained strict data privacy and legal compliance.
• Leadership: Fostered a creative and innovative team environment, delivering a product that elevates the hiring process.

Project: DataOps Platform for MLOps Operations
Role: MLOps Engineer (Lead)
Description:

• Developed a DataOps platform to version ASR and NLP data and manage datasets for an MLOps platform.

Key Achievements:

• Platform Development: Pioneered a DataOps platform integrating both CLI and GUI interfaces for diverse user preferences.
• Cross-Functional Collaboration: Identified and addressed pain points in existing MLOps workflows.
• CLI and GUI Design: Designed a user-friendly CLI for technical users and an intuitive GUI for non-technical users, enhancing model tracking, labeling, and monitoring.
• Version Control: Implemented version control and model tracking features, ensuring traceability and reproducibility of models.
• Training and Documentation: Authored detailed documentation and conducted training sessions for smooth team adoption.

Technologies Leveraged:

• Programming: Python
• CLI Framework: Typer
• GUI Framework: Django, React.js
• Data Processing: Pandas, NumPy
• DevOps Tools: Git, GitHub

Build data pipeline to move data from sources to SnowFlake
Role : Data Engineer
About project : For a Transit Agency data platform, it is required to build ADF pipelines,
To move data from mail as source to Snowflake.
To upload static data files to snowflake
Contribution : Built ADF automation to pull data from Mail and static file and upload to snowflake. Built separate data
copy pipelines for this.
Tech toolset : Azure Data Factory, Azure Delta Lake v2, Snowflake

Build data pipeline to move data from SQLServer to Snowflake in Finance Project
Role : Data Engineer, Bi Developer
About project : A micro finance corporation where their backend is implemented with SQLServer, Salesforce and
Hubspot.
Contributed in movement of data from SQLServer to snowflake. Built ADF pipelines for this, setup alerts for these
pipelines.
Also created interactive tableau reports for financial departments. Built snowflake queries for reports and integrated
with tableau.
Tech toolset : ADF, Snowflake, Tableau

Build data pipeline to move data from SQLServer to SQLServer
Role : Data Engineer
About project : Implement a POC on transformations of SQLServer Tables and store them
Contributed in Transformation of data from SQLServer to SQLServer. Built ADF pipelines for this, setup alerts for
these pipelines.
Tech toolset : ADF