Vision Intel – AI Customer Service Platform

Comprehensive social media management to enhance brand engagement.

Vision Intel (Help+) | Building a Multi-Role AI-Powered Customer Support Ecosystem at Builder.ai

During my nine months at Builder.ai between Q3 2024 and Q1 2025, I served as the Senior Product Manager responsible for transforming Vision Intel (Help+), one of the most ambitious and technically complex products in the company’s portfolio.

Vision Intel set out to solve a challenge no traditional support system could handle: unify customers, merchants, manufacturers, government entities, and super admins under one intelligent platform and enhance the entire ecosystem with advanced AI assistance, AI execution, AI analytics, OCR automation, and multi-role governance.

The product was massive in scope. Five user types, hundreds of workflows, deep operational complexity, and dozens of cross-team interactions. My responsibility was to bring clarity to this complexity: define every flow, structure the entire ecosystem, write all user stories, guide the design and engineering teams, and shape the intelligence layer that would sit at the core of Vision Intel.

Across these months, I built one of the most comprehensive product architectures of my career centered around three advanced AI systems that elevated Vision Intel far beyond a conventional ticketing or support platform.

‍

Designing the Foundation of a Multi-Role Support Ecosystem

Before building the AI layer, Vision Intel needed a strong structural backbone.

I redesigned the product experience across all five roles:

Customers using the mobile app to submit tickets, scan documents, renew warranties, and chat with AI
Merchants resolving customer issues, analyzing complaints, and managing their operations
Manufacturers handling escalations and warranty problems
Government Entities overseeing compliance and regulated service complaints
Super Admins coordinating the entire ecosystem, permissions, analytics, and workflows

Each role had its own purpose, data views, capabilities, and responsibilities and each required its own interface, permissions structure, and user journey.

I rebuilt every part of the product into clear modules: ticketing, OCR, communications, analytics, case routing, escalation layers, SLA enforcement, and customer history. Once the ecosystem was structurally sound, I moved toward the part that would define Vision Intel’s uniqueness: its AI-powered intelligence layer.

‍

The Three AI Systems That Powered Vision Intel

‍

Vision Intel was not “an app with AI.”

It was a full customer-support ecosystem designed around AI.

The platform featured three major AI systems each serving a different audience and purpose. Together, they reshaped how merchants make decisions, how customers get help, and how support teams operate.

The most advanced AI in the entire ecosystem was the Merchant AI Assistant a system designed to behave like an internal data analyst capable of answering complex operational questions, generating dashboards, and transforming raw data into actionable insights.

Merchants normally struggle with navigating mountains of complaint data, service metrics, SLA performance, product breakdowns, and multi-branch activity. The AI system solved this by allowing merchants to “talk to their data.”

With a single message like:

“Show me all unresolved delivery complaints from Riyadh last week, grouped by branch.”

the AI would:

Understand the query
Build a structured data request
Map it to Vision Intel’s database schema
Retrieve the relevant information
Generate a custom-built dashboard
Provide a human-readable summary of insights
Offer follow-up suggestions or deeper analysis

This created a transformative merchant experience. Instead of navigating a complex reporting interface, merchants interacted with their entire operational dataset through a single AI interface.

‍

1. AI Assistant for Merchants: Dynamic Dashboards, Instant Reports & Data Query Engine

With a single message like:

“Show me all unresolved delivery complaints from Riyadh last week, grouped by branch.”

the AI would:

Understand the query
Build a structured data request
Map it to Vision Intel’s database schema
Retrieve the relevant information
Generate a custom-built dashboard
Provide a human-readable summary of insights
Offer follow-up suggestions or deeper analysis

This created a transformative merchant experience. Instead of navigating a complex reporting interface, merchants interacted with their entire operational dataset through a single AI interface.

‍

2. AI Acting on Behalf of the End User: The Action Execution Engine

This second AI system turned Vision Intel into an agent capable of performing support actions on behalf of customers.

Instead of manually filling forms or navigating menus, customers could simply tell the AI:

“I want to file a complaint about Merchant X.”
“Help me renew my warranty.”
“Book a maintenance appointment for next Tuesday.”
“I need to report a damaged product.”

The AI would:

Detect the user’s intent
Fetch all relevant context: user details, past tickets, linked products
Build a draft action (ticket, appointment, renewal request, etc.)
Present a full preview for the user to review
Submit the action upon approval
Track it and provide real-time follow-up through chat

This made the system radically more accessible especially for users unfamiliar with technical forms.

‍

3. Conversational AI for All Users | Multi-Role Knowledge, Guidance & Troubleshooting

The third AI layer was the cross-ecosystem conversational assistant embedded directly in the Help+ mobile app and web dashboards.

It served every user type: customers, merchants, manufacturers, government officials, and super admins, each receiving answers tailored to their context.

The AI could:

Answer questions based on system knowledge
Explain rules, SLAs, warranties, and workflows
Locate tickets, statuses, and histories
Provide troubleshooting advice
Guide users through next steps
Offer relevant actions (“Would you like me to escalate this?”)
Pull structured data from backend systems
Route unresolved queries to human agents with all context attached

This conversational AI became the front door of Vision Intel, ensuring that every user had instant access to guidance without relying on traditional support processes.

‍

Building the Complete Product Blueprint

Beyond AI, Vision Intel required a detailed product architecture.

I rebuilt every core module:

Ticketing
Escalation & routing
OCR document processing
User profiles
Merchant & manufacturer dashboards
Government entity oversight
Notifications & SLA events
Analytics & activity logs
Case lifecycle management
Permissions & multi-role governance

Every part of the product was supported by hundreds of INVEST-style user stories, each with:

Gherkin acceptance criteria
Positive & negative scenarios
Edge cases
Required inputs
Step-by-step flows
Arabic & English messages
Screen-level interaction details

This made Vision Intel one of the most complete, well-documented products I managed and one of the clearest for design and engineering teams to build.

‍

Integrating AI Into the Core Product Experience

What made Vision Intel unique was not that it had AI features, it was that AI reshaped the entire product.

AI didn’t sit on top of the experience.

It was the experience.

Merchants gained a data analyst.
Customers gained an action assistant.
All users gained a conversational support interface.
Documents were processed automatically through OCR + verification flows.
Routing and escalation used AI signals.
Insights were generated automatically.

The product evolved from a traditional support system into a living, intelligent ecosystem.

Impact & Legacy of the Project

During my nine months leading Vision Intel:

I built the entire AI product architecture
Designed all flows for five user roles
Documented hundreds of user stories
Connected design + engineering into one clear blueprint
Reshaped the ticketing & support model around intelligence
Implemented AI that could analyze, act, and communicate
Streamlined the entire ecosystem into a cohesive experience

Vision Intel became one of the most structured, scalable, and innovative platforms within Builder.ai, a true demonstration of how AI can reinvent customer support at scale.

‍