Abraham – Python, AWS, Docker

Real-time civic healthcare feedback platform. It collects voice-based feedback from the public and turns it into structured reports for NHS and local council stakeholders. Features include live AI voice conversations, sentiment-based safeguarding alerts, and an offline kiosk mode for in-person collection.

• Built a bidirectional WebSocket audio pipeline for real-time voice conversations.
• Integrated a multi-model Gemini pipeline for live conversation, transcription and multimodal analysis.
• Designed the MongoDB data layer and the core backend architecture as the primary technical contributor.
• Implemented threshold-based safeguarding alerts that escalated to human reviewers when critical sentiment was detected.
• Built role-based access control with audit-trail logging for logins, updates and deletions.
• Automated document generation for NHS and council reporting requirements.
• Built an ambassador portal with QR-code badge verification, plus an offline kiosk mode with local caching and sync.
• Set up CI/CD with GitHub Actions, deploying to an Oracle Cloud VM via PM2 and Nginx on every push to main.

Real-time monitoring and alerting platform that gives teams operational visibility into production applications. It tracks application health, aggregates logs and fires alerts for live Node.js deployments.

• Designed the platform architecture from scratch and co-built its monitoring, logging and alerting modules.
• Built health-check and log-collection integrations for live Node.js production deployments.
• Implemented alerting rules that surface incidents across multiple production applications.
• Owned backend development, deployment and day-to-day operation of the platform.
• Rolled it out as the monitoring layer for the other production apps he owns.

Wellness platform with mood tracking and a directory of care providers. It was built from scratch and later handed over to an external development agency.

• Designed the data model and API architecture, and built the core backend.
• Implemented the mood-tracking and care-provider directory features.
• Containerised the app with Docker and set up GitHub Actions deployments to production.
• Audited the codebase and wrote architecture and operations documentation for the handover.
• Led the formal technical handover to a third-party agency and supported the transition.

AI-powered haircare personalisation platform with an integrated LLM component running in production.

• Owned production infrastructure on AWS: EKS clusters, ECR registries and IAM.
• Managed containerised deployments on Kubernetes (EKS), including the infrastructure behind the production LLM component.
• Built a GitHub Actions pipeline that automatically built images, pushed them to ECR and deployed to EKS on every merge to main.
• Wrote deployment automation scripts that cut manual release steps and made releases more consistent.
• Configured least-privilege IAM roles and policies for application services.
• Contributed frontend work when needed alongside DevOps ownership.

Project-based DevOps and cloud education platform. The work covered internal infrastructure supporting its live systems.

• Built and maintained containerised deployments with Docker and Kubernetes on DigitalOcean.
• Provisioned and managed infrastructure as code with Terraform.
• Built CI/CD pipelines automating build, test and deployment workflows.
• Maintained Linux-based infrastructure configurations for internal production systems.