2019 - 2023
Bachelor of Computer Science
AMA Computer College
James Raven Tabag
AI Application & Automation Developer
James Raven Tabag
I build practical AI-powered applications, automations, and full-stack web products—from voice agents and support tools to database-backed applications and dashboards.
About Me
Hello, World!
Hey there! I’m Raven, a Computer Science graduate focused on building practical AI-powered applications, workflow automations, and full-stack web products. Before moving deeper into development, I worked in customer support for 1 year and 7 months, where I learned to communicate clearly, understand user frustrations, and approach problems from the user’s perspective. Through self-directed projects, I have developed a browser-based AI voice booking agent and an AI-assisted helpdesk, working across requirements, API integrations, databases, testing, and deployment. I use modern AI development tools to support my workflow while taking responsibility for technical decisions, validation, and the quality of the finished product.
Education
Education
Working method
Process
- 01
Define
Define the problem, audience, constraints, and success criteria before committing to a solution.
- 02
Design
Design the system and interface around clear user flows, resilient data boundaries, and deliberate tradeoffs.
- 03
Develop
Develop in small, testable increments with maintainable code, fast feedback, and security built into each layer.
- 04
Deliver
Deliver a verified release with clear documentation, operational readiness, and a practical path for iteration.
Case files
Projects
GlowNGo - AI Voice Agent
Built a browser-based voice receptionist for a fictional LA salon using ElevenLabs and n8n. Customers can speak naturally with Gigi to book, reschedule, or cancel appointments, while the agent checks live availability and answers questions from the salon’s knowledge base.
- Conversational booking — book, reschedule, or cancel appointments entirely through voice
- Live availability — checks stylist schedules and service durations before offering time slots
- Knowledge-based answers — handles questions about services, pricing, branches, hours, and policies
- Automated workflows — n8n manages booking operations while keeping backend credentials out of the frontend
- Next.js
- React
- Typescript
- Tailwind CSS
- Elevenlabs
- n8n
- Google Compute Engine
- Docker Compose
Daily Plate - AI Assisted Helpdesk
Built a full-stack AI-assisted helpdesk that turns customer emails into threaded support tickets, automatically classifies incoming requests, resolves eligible policy questions, and routes everything else to human agents.
- Email-to-ticket automation — incoming customer emails become organized, threaded tickets automatically
- AI triage — classifies each ticket by category and priority as soon as it arrives
- Safe auto-resolution — handles eligible policy-based questions while escalating everything else to an agent
- Agent assistance — provides AI-generated summaries and one-click draft polishing
- Automatic reopening — customer replies to resolved tickets return them to the active queue
- Vite
- React
- Typescript
- Tailwind CSS
- PostgreSQL
- Render
- Neon
- Resend
- BetterAuth
- Bun
- n8n
- Google Compute Engine
- Docker Compose
Techstack
Skills
Programming Languages
- Python
- Javascript
- Typescript
- SQL
Frontend & UI
- Next.js
- Electron
- React
- Tailwind CSS
- Vite
Backend & APIs
- Express
- Bun
- Axios
- Zod
- Svix
- Resend
Data, Auth & Storage
- SQL
- PostgreSQL
- Prisma
- Neon
- Supabase
- BetterAuth
AI & Automation
- Antrophic
- OpenAI
- Elevenlabs
- n8n
- Prompt Engineering
- RAG
- WebRTC
Testing, DevOps & Monitoring
- Playwright
- BDD
- Vitest
- Allure
- Sentry
- axe
- Docker Compose
- CI/CD
- Wrangler
- Cloudflare
- Render
- Railway
- Google Compute Engine
- Caddy