Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Masters graduate from UC San Diego with production experience building multi-agent LLM systems, RAG pipelines, and AI evaluation frameworks. I’ve delivered end-to-end AI systems—from data pipelines and embeddings to agentic deployment, human-in-the-loop oversight, and responsible AI guardrails. Previously, as a Software Development Engineer at Oracle, I built scalable, cloud-native backend and observability solutions, including event-driven automation and cross-database ETL/CDC infrastructure. I enjoy turning complex requirements into reliable systems that perform in real-world production environments.…Masters graduate from UC San Diego with production experience building multi-agent LLM systems, RAG pipelines, and AI evaluation frameworks. I’ve delivered end-to-end AI systems—from data pipelines and embeddings to agentic deployment, human-in-the-loop oversight, and responsible AI guardrails. Previously, as a Software Development Engineer at Oracle, I built scalable, cloud-native backend and observability solutions, including event-driven automation and cross-database ETL/CDC infrastructure. I enjoy turning complex requirements into reliable systems that perform in real-world production environments.

Ketki Patankar

AI Engineer, Software Engineer, Data Engineer, +3





Masters graduate from UC San Diego with production experience building multi-agent LLM systems, RAG pipelines, and AI evaluation frameworks. I’ve delivered end-to-end AI systems—from data pipelines and embeddings to agentic deployment, human-in-the-loop oversight, and responsible AI guardrails. Previously, as a Software Development Engineer at Oracle, I built scalable, cloud-native backend and observability solutions, including event-driven automation and cross-database ETL/CDC infrastructure. I enjoy turning complex requirements into reliable systems that perform in real-world production environments.…Masters graduate from UC San Diego with production experience building multi-agent LLM systems, RAG pipelines, and AI evaluation frameworks. I’ve delivered end-to-end AI systems—from data pipelines and embeddings to agentic deployment, human-in-the-loop oversight, and responsible AI guardrails. Previously, as a Software Development Engineer at Oracle, I built scalable, cloud-native backend and observability solutions, including event-driven automation and cross-database ETL/CDC infrastructure. I enjoy turning complex requirements into reliable systems that perform in real-world production environments.

Available to hire

Masters graduate from UC San Diego with production experience building multi-agent LLM systems, RAG pipelines, and AI evaluation frameworks. I’ve delivered end-to-end AI systems—from data pipelines and embeddings to agentic deployment, human-in-the-loop oversight, and responsible AI guardrails.

Previously, as a Software Development Engineer at Oracle, I built scalable, cloud-native backend and observability solutions, including event-driven automation and cross-database ETL/CDC infrastructure. I enjoy turning complex requirements into reliable systems that perform in real-world production environments.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Work Experience

Research Assistant at Lo Labs at UC San Diego

March 1, 2025 - Present

Built and integrated the core image analysis pipeline for automated lateral flow assay (LFA) diagnostic strip analysis. Deployed as part of a serverless AWS platform to reduce end-to-end diagnostic processing to under 10 seconds. Developed an automated preprocessing pipeline using a YOLO OBB fine-tuned on 150+ self-labeled strip images and EasyOCR for strip localization, perspective correction, and test label detection under uncontrolled imaging conditions. Worked with AWS Lambda, S3, DynamoDB, API Gateway, and CloudWatch, with deployment managed via AWS CDK infrastructure-as-code.

Research Assistant (Summer Research Intern) at UC San Diego — Lo Labs

March 1, 2025 - Present

Built and integrated a production image analysis pipeline for automated lateral flow assay (LFA) diagnostic strip analysis, deployed on a serverless AWS platform to reduce end-to-end diagnostic processing to under 10 seconds. Developed an automated preprocessing pipeline using YOLO OBB fine-tuned on 150+ self-labeled strip images and EasyOCR for strip localization, perspective correction, and test label detection under uncontrolled real-world imaging conditions. Worked within AWS Lambda/S3/DynamoDB/API Gateway/CloudWatch infrastructure, with deployment managed via AWS CDK (Infrastructure-as-Code).

Software Development Engineer at Oracle Corporation

July 1, 2022 - April 1, 2024

Engineered a metadata-driven ETL migration platform at Oracle Financial Services using Oracle Data Integrator (ODI) to automate cross-database migrations, reducing manual migration effort by 70%. Built timestamp/trigger-based incremental CDC pipelines synchronizing only changed records from Microsoft SQL Server to Oracle with SQL reconciliation to ensure reliable zero-loss migrations. Also developed an event-driven Java TestNG listener to automatically stream 50+ daily UI test executions into Splunk, saving 1–2 engineering hours daily by eliminating manual log collection across DEV, UAT, and PROD. Created reusable SQL/PLSQL log-processing pipelines with regex parsing to transform application logs into structured metrics for production dashboards.

Applied Machine Learning Researcher at Cardiac Design Labs

August 1, 2021 - June 1, 2022

Led a 3-person team building an ML pipeline for non-invasive hemoglobin estimation from PPG signals across 50 patients. Achieved 87.3% accuracy with Support Vector Regression (SVR), 83.5% with Decision Trees, and 93% with Random Forest.