Founder-built AI SaaS
Zellio — AI Commerce Support and Engagement Platform
I founded Zellio and led the product and engineering work from architecture through live operation, building a bilingual SaaS that connects customer conversations, store data, campaigns, and AI-assisted support workflows.
- Role
- Founder and Product Engineer / Technical Lead
- Period
- 2025–present
Founder-built product system
Commerce support, connected
Customer conversations
Store and order data
Zellio platform
AI-assisted workflows
Context
Zellio LLC
Role
Founder and Product Engineer / Technical Lead
Product type
Founder-built AI SaaS
Core capability
TypeScript
Product context
Bringing fragmented commerce support into one product
Commerce support work is spread across messaging channels, store systems, customer records, campaigns, and manual follow-up. Zellio needed to bring those moving parts into a multi-tenant product while remaining reliable against real commerce and messaging APIs.
Attribution anchor
What I was responsible for
As founder and technical lead, I shaped the product and architecture and implemented core work across the web applications, APIs, AI workflows, integrations, mobile client, testing, infrastructure, and production operations. I led and implemented the platform; I do not present it as a one-person codebase.
Selected scope
Selected contributions
- 01
Built unified support and conversation workflows across WhatsApp, Instagram, Messenger, email, Telegram, and onsite chat.
- 02
Delivered Shopify, WooCommerce, Zid, and Salla connections covering synchronization, webhooks, onboarding, and order workflows.
- 03
Implemented AI-assisted knowledge and agent workflows, retrieval, prompt evaluation, observability, document handling, and visual workflow automation.
- 04
Built campaigns and retention workflows with delivery, retry, attribution, recipient inspection, and operational controls.
- 05
Developed authentication, subscriptions, analytics, internal operations, bilingual English/Arabic experiences, and an Expo support application.
- 06
Established repeatable cloud infrastructure, deployment pipelines, background processing, testing, monitoring, recovery practices, and production incident handling.
Evidence, not reconstruction
System view
A simplified view of verified responsibilities and system relationships. It does not reproduce a private or historical interface.
Founder-built product system
Commerce support, connected
A simplified view of the product boundaries I led and implemented.
Customer conversations
Messaging, email, Telegram, and onsite chat entry points.
Input
Store and order data
Commerce connections, synchronization, webhooks, and lifecycle events.
Input
Zellio platform
Multi-tenant product, permissions, operations, subscriptions, and analytics.
System
AI-assisted workflows
Knowledge, retrieval, agents, evaluation, and reviewable automation.
Process
Campaigns and retention
Delivery, retries, attribution, and operational controls.
Process
Web and mobile operation
Bilingual experiences for support teams and internal operators.
Output
Verified relationships
- Customer conversationsZellio platform · conversation events
- Store and order dataZellio platform · commerce events
- Zellio platformAI-assisted workflows · context and controls
- AI-assisted workflowsWeb and mobile operation · assisted handling
- Zellio platformCampaigns and retention · audiences and events
Engineering judgment
Selected technical decisions
Decision 01
Used a TypeScript pnpm/Turborepo structure to share product, authentication, and integration code across Express, Next.js, React, and Expo applications.
Decision 02
Combined PostgreSQL with Prisma and Drizzle, Redis-backed processing, BullMQ, Inngest, and event/webhook flows for long-running and failure-prone work.
Decision 03
Integrated multiple AI providers and workflow tooling, while managing AWS and Cloudflare infrastructure through Terraform and staged delivery practices.
Technology context
- TypeScript
- Next.js
- React
- Express
- Expo
- PostgreSQL
- Prisma
- Drizzle
- Redis
- BullMQ
- Inngest
- Mastra
- Terraform
- AWS
- Cloudflare
- Playwright
- Vitest
Delivered capability
“A live bilingual product across web and mobile, demonstrating end-to-end ownership from ambiguous product idea through architecture, integrations, testing, deployment, and continued production improvement.”
AI-enabled SaaS · Commerce integrations · Product engineering