Kelvin Emwinghare
I am Kelvin Emwinghare, a Senior Data Scientist with 6+ years building production ML systems, data pipelines, and analytics platforms.
About
I am Kelvin Emwinghare, a Senior Data Scientist with 6+ years building production ML systems, data pipelines, and analytics platforms.
I specialize in credit risk modeling, fraud detection, and product analytics, with hands-on experience in Snowflake, dbt, Meltano, Prefect, and real-time scoring architectures. I have delivered end-to-end solutions across fintech and consulting, including RAG architectures and cross-jurisdiction AML compliance, driving measurable business impact.
- 6 years' experience
- Joined October 2024
Industry experience
- Financial Services
- Software & Internet
- Professional Services
- Other
Work experience· 7 roles
Built a web application for claims administration using JavaScript, Apex, and Visualforce; improved case processing efficiency and reduced manual handling by ~25%. Developed Salesforce reports and dashboards; created a POC LLM-powered analysis system using Azure OpenAI GPT-4 to automate case investigations and improve compliance review accuracy. Designed ETL pipelines moving Salesforce data to Azure databases for integration with Microsoft Fabric, with incremental loads and data validation, reducing data discrepancies by ~36%.
Led a hybrid LLM and ML-driven fraud detection system for fiat on-ramp AML across multiple jurisdictions, reducing manual reviews by 34% and increasing fraud detection by 27%. Implemented staking reward calculations in dbt for 242k+ users, contributing to 100k+ new users and $5M+ staking AUM. Built a forecast solution with XGBoost and rule-based logic to predict weekly withdrawal volumes, improving liquidity and revenue by $300k+. Designed and maintained Omni and Mode dashboards for fiat on-ramp, staking, and distributions, enabling near real-time visibility. Owned end-to-end data ingestion pipelines with Meltano and Prefect, including custom taps and Snowflake ingestion; automated Terraform provisioning.
Built a web application for claims administration using JavaScript, Apex, and Visualforce, improving case processing efficiency and reducing manual handling by ~25%. Developed Salesforce reports and dashboards to monitor claims administration activities and performance metrics. Built a POC LLM-powered analysis system using Azure OpenAI GPT-4 within Salesforce, automating case investigations and enhancing compliance reviews. Designed ETL pipelines to move Salesforce data to Azure databases for integration with Microsoft Fabric, implementing incremental loads and data validation that reduced data discrepancies by ~36%.
Education
Hire Kelvin Emwinghare today
Post a job and invite Kelvin, or contact them directly. No hidden commission.