ENG - SI Engine

• Low-level RestAPI calls of connection creation on Fivetran, data sync in Databricks and tenant creation in MicroStrategy.

• Databricks Security design: Create catalog using tenant-specific service principal, limit the visibility of customer data; grant USE privilege to admin service principal for management

• Engineered and maintained the backend architecture for a multi-tenant, AWS-based enterprise BI platform, integrating Salesforce, ServiceNow, Excel, and other sources to deliver a unified data retrieval experience;

• Automated tenant provisioning on MicroStrategy and data integration with Fivetran and Databricks using API calls and created restricted service principals on Databricks;

• Developed internal APIs to trigger provisioning processes and implemented Fivetran webhooks to automate data synchronization to Databricks, facilitating dataset and analytical card creation on MicroStrategy with robust retry and cleanup mechanisms;

• Managed version control and updates for datasets and cards by adhering to predefined manifest files, ensuring consistency and accuracy across all tenant environments.

AA-AI Analytics

• Harden searching feature with comprehensive indexing performance testing and concurrency testing using Jenkins and Kubernetes;

• Developed a Python tool utilizing REST API to log the distinct element counts of string-type attributes within an environment, helping engineers determine accurate tokens and resources for indexing;

• Prompt externalization: replace hard coded prompts with YAML files in order to change the prompts used by the AI Service without changing any source code;

• Dynamic script generation: Using GenAI for prediction and visualization, streamlining the process of moving out data from MicroStrategy for predictive analysis on other ML platforms.

SI-Streaming Engine

• RabbitMQ

Skills: AWS CloudFormation, AWS Sagemaker, AWS EC2, Python, React.js

• Migrated the machine learning model for AI-detection from local to AWS cloud, utilizing AWS CloudFormation, EC2, and SageMaker to enhance scalability and performance;

• Developed a web application using React.js and FastAPI enabling image uploads for analysis, ensuring seamless user experience across devices and adherence to WCAG accessibility guidelines;

• Ensured robust data security and infrastructure design, safeguarding sensitive user information and improve system reliability;

• The product reaches over 90% accuracy of detecting the percentage of an image is AI-generated by Stable Diffusion or Pro GAN.

Skills: AWS CloudFormation, AWS Lambda, Java, TypeScript, AWS S3, AWS Athena, CDK pipeline, SQL

• Developed CDK pipeline for record handling using AWS Lambda, S3 buckets, and Athena Database across alpha/beta/prod stages, including integration tests.

• Detected and addressed 2 security vulnerabilities within the first month, ensuring Client-side Encryption for records in production services.

• Mitigated KMS throttling, reducing potential service downtime from 2 hours to zero interruptions.

• Introduced a pre-release impact assessment feature, seamlessly integrated with daily operations.