Job – Node.js, Typescript, React

A multi-tenant SaaS platform that helps U.S. entrepreneurs form and run their businesses, from company registration to ongoing compliance, legal, and tax filings. Its business formation flow is the core customer funnel, and traffic spikes sharply during tax season. Customers rely on support for questions about formation, compliance, and filing that often require real-time lookups of their own account and company data.

• Designed and deployed a custom MCP server with vector search that lets clients autonomously query their own company and filing data, reducing tech support and customer service workload by 30%.
• Drove the migration of the business-critical formation flow from a PHP monolith to Next.js, React, TypeScript, and NestJS, introducing workers, queues, and event-driven patterns, which cut page and request load times by 60% and pushed Core Web Vitals into the top 10th percentile.
• Migrated CI/CD to GitHub Actions with AWS OIDC, reducing deployment time from ~2 hours to ~12 minutes.
• Provisioned the cloud infrastructure with Terraform (CloudFront, S3, RDS Postgres, Route 53, WAF) and Kubernetes on EKS, choosing EKS over ECS to scale through tax-season traffic spikes.
• Built a production AI voice assistant for the support team on RAG pipelines with LangChain and a vector database, including evaluation, observability, and guardrails against hallucinations and legal-advice risk.
• Engineered an internal AI agent framework on Claude Code and Antigravity (parallel agents, custom skills, hooks, roles, and commands) to automate engineering workflows.
• Introduced FinOps cost-allocation tagging and spend alerts to catch runaway cloud costs; led two mid/junior teams and mentored engineers to increase shipping velocity.

A video management module of a cloud-native digital publishing platform used by major news and media organizations. It ingests, transcodes, packages, and delivers live and on-demand video at scale, including server-side ad insertion, by orchestrating a pipeline of AWS media services through event streams.

• Debugged and extended the Express microservices handling video ingestion at scale for the platform's video pipeline.
• Integrated Kafka event streams in Go to orchestrate AWS Media Services (MediaLive, MediaStore, MediaTailor, MediaPackage, MediaConvert) across live and on-demand workflows.
• Managed containerized deployments on Kubernetes clusters via Helm charts.
• Delivered this engagement in parallel with a second Wizeline client project, contributing at a senior level on both.

An identity security platform that brings single sign-on and automated access management to SaaS applications that lack native SSO support. It uses browser automation to log in, rotate credentials, and provision or deprovision users on behalf of enterprise IT and security teams, closing a common security gap. The automation layer runs on a proprietary TypeScript framework shared by several engineering teams.

• Refactored the proprietary TypeScript framework underpinning the web-scraping SSO platform, reducing bundle size by 40% and enabling independent deployments across five engineering teams.
• Introduced Docker-based local development environments, cutting engineer onboarding from days to hours.
• Led development of new platform features on NestJS and TypeScript while the framework refactor was in flight.
• Delivered senior-level contributions on this engagement in parallel with a second Wizeline client project.

An enterprise video platform used by media companies and Fortune 500 brands to host, stream, monetize, and analyze video. Enterprise prospects evaluate it through fully customizable, live demo environments that show adaptive streaming, DRM-protected playback, and player plugins in action. The platform's web player is built on Video.js, a widely used open-source video player.

• Designed and built production-grade, fully customizable demos used live in enterprise sales cycles with Fortune 500 clients, directly supporting approximately 70% of deal closures at the negotiation stage.
• Contributed to Video.js, working on adaptive bitrate logic, buffer management, and the plugin architecture.
• Showcased the platform at its technical ceiling across HLS, DASH, HEVC, and DRM-protected streaming.
• Built the demo platform on the MERN stack with a real-time signaling layer, deployed on AWS (EC2, ECS, Lambda, API Gateway, ELB, CloudFront, Elasticsearch) with all infrastructure provisioned in Terraform.
• Containerized every service with Docker to guarantee environment parity and eliminate failures during live client presentations.

A mobile application for a U.S.-based dental clinic business, backed by an API and geo-distributed databases. The setup keeps the app fast and available for users across different locations, with the backend deployed on auto-scaling cloud infrastructure.

• Built the React Native mobile app and the NestJS API behind it end-to-end.
• Designed geo-distributed PostgreSQL instances to optimize latency and availability for users across locations.
• Deployed the backend with Docker on AWS EC2 with auto-scaling groups.

An automated billing system for a U.S.-based business that runs high-volume, transaction-critical charging and invoicing workflows. The engine processes up to 300,000 transactions per day and is built for fault tolerance and horizontal scalability, so billing keeps running through load spikes and partial failures. An internal admin console gives operations teams tools for billing reconciliation and day-to-day operational tasks.

• Architected and built the high-throughput billing engine in Node.js end-to-end, processing up to 300,000 transactions per day at 99.9% uptime.
• Led the system design around idempotency, sharding, elastic indexing, fault tolerance, and horizontal scalability.
• Designed the data and messaging layer with PostgreSQL read replicas, Redis caching, and Kafka for event-driven workflow orchestration.
• Built the internal NestJS admin layer for billing reconciliation and operational tooling.
• Deployed the system with Docker on AWS EC2 with auto-scaling groups.