Michael Pereira

Led design and implementation of a Multi-Vector Index architecture for an enterprise AI platform, enabling multi-embedding RAG workflows and improving retrieval relevance by 30% while increasing stability via vector index guardrails.

Built an end-to-end AI response evaluation system using watsonx.data, integrating truthfulness, reliability, and cost-performance metrics via APIs, persistence schemas, and UI integrations to enable data-driven optimization of AI outputs.

Integrated Langflow into a Kubernetes-based AI platform by redesigning containerization and CI/CD workflows with optimized multi-stage Docker images, hardened health checks, and automated deployments, improving release reliability and reducing deployment overhead by 40%.

Optimized AI inference by improving SSE and chunked streaming across frontend and backend systems, enabling incremental response rendering that reduced perceived latency by 35% and resolved streaming reliability issues.

Designed and implemented a centralized MCP Gateway for all MCP servers, unifying authentication, authorization, and logging into a single entry point, reducing integration complexity by 30% and improving observability and governance.

Developed Angular/TypeScript features for payments and travel platforms, optimizing critical customer journeys and reducing task completion times by 20% through improved workflow design and usability.

Improved performance and responsiveness of accounts and transactions micro-apps by optimizing UI behavior, enhancing mobile responsiveness, and strengthening Jest-based test coverage, while improving CI/CD reliability via Bitbucket and Jenkins, reducing page load times by 50%.

Architected a scalable frontend component system that improved rendering performance and standardized WCAG-compliant accessibility through ARIA, keyboard navigation, and semantic HTML.

Collaborated with leadership, product owners, and cross-functional Agile teams to translate business requirements into scalable technical solutions, improving delivery alignment and reducing downstream rework and ambiguity.

Owned production support for core banking services, resolving high-severity incidents, remediating critical vulnerabilities, and improving accessibility to enhance platform reliability and compliance.

Created a React, TypeScript, and Electron control application for a custom 6-DOF MRI-compatible robotic phantom, automating Arduino-driven motion sequencing and supporting sub-millimeter motion validation while reducing manual testing effort by 40%.

Designed and implemented a scalable full-stack medical imaging platform using React, Node.js, Redis, and AWS, integrating AI-driven DICOM processing pipelines that automated fetal MRI segmentation, motion correction, and anatomical classification, standardizing clinical evaluation workflows, and reducing dependence on specialized radiology expertise.

Engineered a scalable cloud-based surgical assessment platform using Nuxt, Spring Boot, and AWS (S3, DynamoDB, API Gateway, EC2), architecting media processing and storage workflows that improved upload performance by 25% and supported growth in clinical assessment volume.

Launched a Google Chrome extension that creates an advanced Google event and provides additional in-meeting features. This extension was made using HTML, CSS, JS and Bulma.

Built a dynamic dashboard with React, TypeScript, Tailwind, and Chart.js, providing a user-friendly interface for client analytics.

Developed middleware with TypeScript, Express, and Redis to integrate Airtable APIs, streamlining data access.

Collaborated directly with clients to gather requirements, address feedback, and provide project updates, ensuring high satisfaction and seamless delivery.

Utilized project management tools like Jira to prioritize tasks, coordinate with the team, and ensure timely dashboard deployment.

Developed a client application leveraging React Native, TypeScript, and IBM's Carbon Design System with Cognos Analytics services.

Collaborated with cloud and full-stack developers, QA engineers, and designers to deliver features and enhancements.

Released high-quality updates with bug fixes and new features for both Apple's App Store and Google's Play Store.

Led the revamp of e2e test structures and spec files, improving testing efficiency and coverage.

Contributed to the maintenance and development of new features for an accounting software suite, leveraging full-stack technologies including Angular, TypeScript, Spring, Java, C#, and AWS.

Created and executed migration scripts for data handling with MSSQL and MongoDB.

Designed effective user interfaces with CaseWare's Nova CSS framework, enhancing usability and aesthetics.

Consistently exceeded sprint deadlines for bug fixes and features, earning client and stakeholder recognition.

Used Java, TestNG, and Gradle, to write API tests to verify the back-end server's logic and its API requests/responses.

Worked with TypeScript and Cypress, to write front-end tests to verify elements on the page and to ensure intended client usage/behaviour is working accordingly.

Obtained a profound understanding of scalability, deployment procedures, and microservice architecture within a monolithic repository.

Excelled using AWS resources for product features and testing such as Dynamo, SQS, SNS, SES, API Gateway, and Lambda functions.

Created GitHub workflows and actions for CI to run tests on PR and Post Deployment.

Hosted agile meetings such as daily stand-up and sprint retrospectives, contributed to weekly design and architecture meetings and participated in bi-weekly demos.