Founder-led full-stack AI consultancy in Vancouver — I build RAG chatbots, agentic AI workflows, and AI automation for SMBs, from architecture to launch. No handoffs.
TypeScript · Next.js · AWS · Claude API · Playwright
Ask Miko about SI8
Don’t take my word for it — the chatbot answering you is the kind of system I build. Ask it anything about my services, process, or pricing.
See pricing & product details →Aggregated across prior roles and projects — detailed in the work below.
Every engagement is scoped to outcomes — not hours billed.
You get automated pipelines that replace hours of manual work daily. I integrate LLMs, build multi-agent orchestration, and wire AI into your existing systems so your team ships faster.
You get a production-ready web application — from database design to deployed UI — built with modern TypeScript, React, and Node.js without the overhead of a large team.
You get a clear architectural roadmap before a single line is written. I map business requirements to technical decisions and flag risks before they become expensive.
You get a working modern system without losing years of business logic. I migrate legacy codebases to current stacks incrementally — no big-bang rewrites.
You get infrastructure that scales with demand and costs less to run. I design and implement AWS-native architectures with CDK, Lambda, and proper observability.
You get confidence in your release pipeline. I build Playwright and Selenium test suites that catch regressions before they reach production — 30% faster release cycles.
Products I've designed, built, and launched end-to-end — under SI8 and independently. Not résumé bullet points: live software you can open right now.
Guardrailed multi-tenant RAG chatbot — shipped as SaaS
A productized retrieval-augmented chatbot platform. Supabase pgvector retrieval, a deterministic eval gate that blocks deploys on information leaks, durable cross-instance rate limiting, and Stripe subscription billing with webhook-driven tenant activation. It runs live on this very site.
AI-powered personal finance — no spreadsheets, no manual entry
Connects to bank accounts via Plaid, auto-categorizes transactions with AI, and surfaces insights — a life-cost view (spending shown as hours of working life), 20-year net-worth projections, and a chat assistant that answers questions about your money.
Sunset-quality forecasting for photographers
Scores sunset and golden-hour quality worldwide from weather and atmospheric conditions, with location guides and a live planner — so photographers can plan the shot before the sky changes.
Browser-native, repeatable AI prompt workflows
A Chrome extension for building, saving, and running multi-step AI prompt chains with dynamic input templates and reusable collections. Privacy-first — data stays local unless explicitly synced.
This site runs my own retrieval-augmented (RAG) chatbot, “Miko” — I built and deployed it end-to-end. Ask it whether SI8 is a fit for your project: it answers from a knowledge base about my services, process, and background.
The bot you’re talking to is the kind of system I can build for you.
SI8 Technology launched in 2025. The work below was delivered during my prior roles — it represents the caliber of engineering I bring to every engagement.
Each entry notes whether it was personal work or done as part of an employment role.
Problem
I set out to test whether complex multi-vendor purchasing workflows could be fully automated. The target workflow involved 4–6 hours of daily manual work across multiple vendor portals — error-prone and unscalable.
Approach
Built a multi-step Playwright automation pipeline triggered by incoming purchase orders. Integrated validation logic and exception handling to catch edge cases without human review.
Result
Achieved 300% faster cycle time in controlled testing vs. manual operation. The automation handles the full workflow end-to-end. Production deployment in progress.
Problem
A growing SaaS platform needed new data-driven features and improved reliability to serve a rapidly expanding user base across North America.
Approach
Full-stack feature development in Vue.js / Django. Built an AI-driven content recommendation system that personalized the discovery feed per user segment.
Result
Platform reached 5,000+ active organizations. 80% reduction in manual data entry. 99.9% uptime maintained, with mean time to recovery improved 65% through Datadog monitoring and observability.
Problem
A growing engineering team needed to ship faster without sacrificing code quality, with AWS infrastructure costs climbing as product demand scaled.
Approach
Introduced agentic AI tooling (Claude Code, Cursor) into the core development workflow. Refactored AWS Lambda architecture with CDK-managed infrastructure.
Result
2–3× increase in feature delivery velocity — roughly 30% shorter development cycles plus parallel agent workflows that multiplied throughput. 30% reduction in AWS infrastructure costs.
Peer recommendations from engineers I've worked with directly.
Software Engineer @ DMG Blockchain
"SI8 Technology possesses an innate ability to truly understand the essence of a product, ensuring that every feature and functionality aligns seamlessly with the end-user's needs and the overarching business objectives."
Full Stack Software Engineer @ GrainFox
"If you're looking for a highly skilled software engineer who not only delivers results efficiently but also fosters a positive collaborative environment, Michael is the ideal choice."
Every project is led by me — not delegated.
Founder & Principal Engineer
I'm an AI-native full-stack engineer who founded SI8 Technology to offer the kind of senior-level execution that most agencies can't provide — direct collaboration, zero handoffs, and accountability at every step.
With 6+ years across product companies, I bring a systems-first perspective shaped by my background in Mechatronic Systems Engineering. I think in architectures, not just features — and I bring that to every engagement.
I leverage cutting-edge tools to build robust, scalable solutions.
Next.js
React
Vue.js
TypeScript
Python
Node.js
PostgreSQL
Supabase
Django
AWS
Tailwind CSS
Playwright
GitHub
Vercel
Datadog
Expo
Straight answers before you reach out.
Tell me about your project. I reply within 2 business days.
Describe your project and I'll get back to you with a clear next step.
I learn your goals and constraints in a focused 30-minute call.
I map your requirements to a technical roadmap and flag risks early.
I build and ship — direct accountability, no handoffs.
I stay available post-launch for maintenance, iterations, and scale.