Distinguished Engineer, Director of Architecture at Cognizant Accelerator Studio (March 2019 – October 2022)
As the Director of Architecture at Cognizant Accelerator Studio, I led a team of 16 architects overseeing 29 products, guiding technical direction, and driving innovation across multiple programs. My tenure involved spearheading significant projects and implementing robust solutions that transformed internal processes and client offerings. Cognizant Accelerator Studio was started as an incubator for internal projects that were selected to go to market as products.
Key Projects:
IPReady InnerSource Program:
- Co-founded and led the program, cataloging over 4100 IP assets.
- Established DevSecOps and “shift left” standards.
- Implemented CI/CD infrastructure with 20,000 GitHub seats and 30,000 GitLab seats.
- Each project was scored and classified using tooling such as SonarQube to identify suitability for delivery or re-use.
- Led development of the CodeHub Asset Catalog web app using React, ExpressJS, and MongoDB.
QuickStart Project Launch:
- Created an internal Heroku-like platform for rapid application deployment on AWS.
- Technologies included React, Angular, Keycloak, Spring Boot, MongoDB, and PostgreSQL.
- Designed and implemented platform DevEx for OAuth2, OIDC, Logging, CI/CD and other NFRs and cross cutting concerns.
Java Transformation Program:
- Developed template reference architectures for Java applications in collaboration with Pivotal Labs.
- Managed a $52,000 weekly budget contracting budget, creating project templates that were later used in hundreds of projects.
- Templates covered aspects of 12-factor applications.
Strategic Partners Program:
- Designed and oversaw projects including Microsoft-based bots, AWS telemedicine chat services, 5G auto provisioning with Verizon, and React/NodeJS POCs.
- Followup to CIQA project using AI/ML to transcribe video chat, identify key terms, and autofill forms on the fly during agent and customer conversations.
Cognizant Insurance Quote Accelerator:
- Led the development of an AI/ML-based system to simplify insurance applications.
- Used MarkLogic Data Hub to build the central address database, leveraging fuzzy matching, Levenshtein distance, and metaphone phonetic matching.
- Worked with teams to build classifier to select the NAICS code based on company name and description.
- Other supervised training using AWS Sagemaker and hand built training examples leveraging Cognizant’s massive employee resources.
- Recruited the Chief Data Scientist and helped to build the AI/ML team.
- Managed data models and ETL to ingest address data from 9 data providers.
- Technologies included React Native, Spring Boot, ArangoDB, AWS Sagemaker and MarkLogic Data Hub.
Consulting Architect for Cognizant Shared Investigator Platform:
- Research root cause for severe bottlenecks in product.
- Presented remediation plan to CTO, CIO.
- Plan allowed for SOLR indexing of billions of documents.
- Discovered and fixed issues with Oracle DB hosted on AWS.
- Presented roadmap to large Pharma customers.
- Technologies included SOLR, Spring Boot, Mulesoft, AWS, on premise VMware.
Key Achievements:
- Delivered innovative solutions that streamlined processes and enhanced client offerings.
- Demonstrated leadership in driving technical strategies and architectural decisions.
- Fostered a collaborative environment, establishing an internal developer’s craft community.
- Successfully led multiple high-impact projects, ensuring they met business and technical objectives.
Tools and Technologies:
- UI Frameworks: React, Angular
- Backend: NodeJS, ExpressJS, Javascript, Typescript, .NET, Java, Spring Boot, Python
- Databases: MongoDB, PostgreSQL, ArangoDB, MarkLogic Data Hub
- Cloud and DevOps: AWS, GitHub, GitLab, Terraform, Kubernetes, RabbitMQ
- Identity Management: Keycloak, Cognito
- AI/ML: AWS Sagemaker
My experience at Cognizant Accelerator Studio showcased my ability to lead large-scale projects, implement cutting-edge technologies, and drive strategic initiatives that deliver significant business value.