Full-Stack AI Engineer
I love creating software that is clean, useful, and easy to use. Lately that has meant moving beyond "AI user" habits and into AI-assisted engineering systems: clarifying requests, shaping specs, carrying product context forward, giving agents the right repo knowledge, and reviewing the output before anything ships.
The code output has gone way up, but the judgment around the code matters more than ever. My focus is the workflow around the work: product intent, data models, architecture decisions, context engineering, review loops, and the rails that make AI-assisted engineering trustworthy.
I still really enjoy creating websites with rich UI components, including:
- Web applications,
- Dashboard layouts,
- CMS contents layout,
- AI workflow tools,
- Documentation and planning systems,
- and anything I can make more beautiful and fun.
But I still like to make simple website pages like landing pages. So, what tools do I feel comfortable using while building products and workflows?
Technologies I Have Experience With
Programming Languages
AI Coding Tools & Agent Workflow Technologies
Front-end & Back-end Technologies
Database Technologies
Supabase (Database / Auth / Storage / Edge Functions / Realtime)
PostgreSQL PrismaFirebase (Hosting / Storage / Auth)
Softwares and Tools
CMS Website Builders
AI & Agent Workflows
I’m a hands-on full-stack engineer first. AI pairing is an accelerator I design and oversee—using curated prompts, repo context, product intent, rules, automation, and review gates to keep quality high without losing architectural direction or UX polish.
If the plan is vague, the data model is wrong, or the agent misses the product intent, AI just produces more wrong code faster. The useful leverage comes from upstream clarity and downstream verification.
Patterns I use: agentic planning, context loading, dedupe, enrichment, implementation, review, QA, status reporting, and automated checks—with human-in-the-loop only where judgment is actually needed.
Current Favourite Tech Stack / Tools
TypeScript is my default for full-stack work. I design type-safe APIs, domain models, and UI contracts so features stay reliable from database to browser.
Next.js (App Router) with React is my primary web stack. I ship production-ready dashboards, multi-tenant role-based access, and real-time collaboration with confident routing, caching, and SSR/ISR patterns.
Favourite UI system: Tailwind CSS paired with shadcn/ui and Radix primitives, orchestrated with TanStack Query for server state and caching. This stack lets me move quickly while still delivering polished, accessible UIs.
Shadcn UI is my go-to for speed of development while being flexible enough to customize heavily. When working with other frameworks—I can customize tokens and theming so it feels bespoke, and I'm comfortable with Material UI, Chakra UI, Hero UI, and other component frameworks.
Framer Motion is my go-to for purposeful motion design—page choreography, micro-interactions, and responsive states that make the UI feel intentional.
Rive adds interactive, state-machine-driven vector animations when I want hero moments, empty states, or playful feedback without compromising performance.
Story Of My Coding Journey
My path blends seven years as a BIM technologist with 6 years of full-stack development.
From BIM to software
- Built 3D BIM models and project workflows, then automated bottlenecks with Dynamo for Revit, PyRevit (IronPython + .NET), and native C# add-ins—turning 4-hour manual tasks into 10–15 second runs and rolling them out company-wide.
Leveling up for the web
- I was introduced to web development and learned JavaScript and TypeScript via Vue.js/Vuetify, then moved into React/Next.js ecosystems. Early projects covered chat apps, game logic, and REST API integrations with MongoDB/Mongoose.
- At InceptionU, led teams through Next.js, Nest.js, Supabase, and testing with Jest, establishing CI/CD and teaching new tech quickly.
Building products at scale
- Collegium (Full Time, Oct 2023 – Jul 2025): was my first full-time role, and was the company's founding developer. Initially delivered a fully functioning MVP for their three-sided real-estate platform, in a short 3-month period, going on to fleshing out the product with more business-oriented features later. Built Firestore data models with real-time listeners, added TanStack Query caching + optimistic updates, designed modular Next.js front-ends with role-based access, and integrated Autodesk APS to ingest 50k+ BIM elements for interactive 3D views to enable automated quantity take-offs and tracking element lifecycle to manage scope.
- Village Wellth (Contract, Aug 2025 – Nov 2025): was a chance to work inside an experienced, tight-knit team that felt like a small powerhouse. The quality bar was set by people who cared deeply about the product and the craft, and it sharpened how I think about building in a real production codebase.
- My contract time there also deepened the way I use AI-assisted coding. I kept learning that good agent output depends on strong repo context, clear prompts, and active in-session guidance. Earlier models could move quickly, but they could still hallucinate details or miss architectural intent, so detailed code review, tests, and human product judgment remained essential.
- Max Technologies / Webvar (Full Time, Dec 2025 – Present): has become the place where AI workflows moved from side experiments into the rhythm of everyday engineering. I have been pushing to embed AI more deeply into planning, development, review, QA, and team communication—not as a replacement for judgment, but as a better operating layer around the work.
- That effort has grown into developing an in-house code factory tailored to the company's products, process, and quality bar. It has pushed me beyond asking whether AI can write code and toward designing reliable engineering loops where context, review, and coordination help the whole team move faster without losing product intent.
What drives me
- I like turning ambiguity into shipped features: mapping workflows, prototyping UX, and then delivering type-safe APIs, polished UIs, and reliable tests. Strong prompts and solid architecture keep AI-assisted workflows aligned with product goals. The magic is in the details, and I have a passion for delivering on high-quality user experiences.