Celigo
AI-Powered Integration Assessment Tool
Strategic AI integration through human-centered design
2025
Celigo had invested in building a custom GPT trained on years of proprietary integration data from thousands of customer implementations. The AI was sophisticated, capable of providing valuable recommendations about complex integration challenges. But there was a fundamental problem: to get meaningful insights, users faced a long, arduous process of typing detailed, unstructured descriptions of their tech stack, integration architecture, and pain points through a basic chat interface. The real challenge wasn't the AI's capability—it was recognizing that sophisticated AI needs the right interface to deliver business value. I designed and built a structured assessment wizard that transformed friction into flow, converting an underutilized chatbot into a strategic tool that captured qualified leads and rich business intelligence for sales enablement.
Roles, Tasks
Strategic Planning, UX/UI Design, Front-End Development (AI-Assisted), System Architecture, Cross-Functional Collaboration
The Challenge
Transform an underutilized custom GPT into a high-value lead magnet that reduces friction for prospects while capturing detailed intelligence that enables sales teams to prioritize leads, optimize conversations, and close deals faster.
The existing custom GPT represented significant investment but minimal business impact—a common pattern when organizations focus on AI capabilities without addressing the interface challenges that prevent those capabilities from being realized.
Marketing had built a custom GPT trained on Celigo's extensive integration expertise, but the default chat interface created massive barriers to value. Users had to manually type lengthy, unstructured descriptions of their current tech stack, integration methods, challenges, and goals. This created multiple problems: the time investment discouraged completion, the unstructured data varied wildly in quality and terminology, and the friction prevented users from providing the comprehensive information needed for truly valuable recommendations. The result was a technically sophisticated AI tool generating minimal business value. Marketing struggled to capture qualified leads beyond basic contact information. Sales received limited intelligence about prospect situations, making lead prioritization difficult and initial conversations less efficient. Most critically, prospects rarely received recommendations compelling enough to justify the effort, undermining Celigo's positioning as an integration thought leader at a critical point in the buyer journey. The gap between AI capability and business impact wasn't a technology problem—it was a design problem.
The strategic solution
The existing custom GPT chatbot: powerful AI trained on proprietary data, but limited by an interface that created massive friction for users.
Rather than rebuilding the AI, I designed a complementary front-end experience that structured data collection, reduced cognitive burden, and transformed the interaction from interrogation into guided assessment.
The solution leveraged another AI tool—Figma Make's AI-assisted development capabilities—but the strategic value came from applying 15+ years of UX design expertise to solve a fundamental interface problem. This wasn't about what AI could do; it was about understanding what users needed to provide, what sales teams needed to receive, and designing the bridge between those requirements.
The assessment wizard approach solved multiple challenges simultaneously. First, it dramatically reduced friction through visual recognition and structured inputs. Instead of typing app names, users selected from recognizable logos organized by functional category (finance, HR, sales, etc.). This wasn't just faster—it eliminated spelling variations, naming inconsistencies, and the cognitive burden of recalling exact product names. Text labels beneath logos provided redundancy, and custom input options ensured flexibility.
Second, it standardized data structure. Predefined categories and consistent terminology replaced free-form descriptions. Integration types became clear options (native, prebuilt, custom API, manual) rather than varied explanations. This structured approach enabled the AI to provide more accurate, specific recommendations while creating clean data for CRM integration.
Third, it incorporated behavioral design principles. Breaking the flow into discrete steps prevented the "mountain of work" perception. Progressive visual feedback—dots appearing in the background for each app added, lines connecting integrated applications—provided positive reinforcement that encouraged comprehensive responses. These weren't decorative elements; they were strategic nudges that increased data quality by making the input process feel rewarding rather than burdensome.
The approach transformed the value exchange: prospects received genuinely personalized recommendations based on their specific situation rather than generic content, while Celigo captured qualified leads with unprecedented technical detail—all structured data flowing directly to the CRM for immediate sales team use.
Process
I developed a multi-stage assessment architecture that balanced user experience requirements with business intelligence objectives, then orchestrated AI development tools to build the complete system.
A human-centered solution
The wizard consisted of four strategically designed stages, each solving specific problems:
Stage 1 - Foundation & Context: Basic lead capture (contact information) paired with firmographic qualification (company size, revenue range, order volume). This immediate qualification data enabled sales to assess ICP fit before deeper technical review.
Stage 2 - Tech Stack Mapping: The core innovation—visual app selection organized by functional process areas. Users clicked recognizable logos rather than typing names, with three levels of redundancy: logo recognition for speed, text labels for clarity, and custom input for edge cases. This approach reduced cognitive load while ensuring data accuracy and consistency across submissions.
Stage 3 - Integration Architecture: Displayed only previously selected apps, then enabled users to map connections and classify integration types. Making this stage optional maintained momentum—users could skip for quick results or provide detailed architecture for sophisticated analysis. The focused display (showing only relevant apps) kept the task manageable even for complex environments.
Stage 4 - AI Processing & Response: Behind the scenes, all captured data exported as structured JSON with a custom prompt, sent to Celigo's GPT for analysis, then returned as formatted recommendations. The results page organized insights by implementation effort and business priority, naturally incorporated relevant Celigo capabilities, and concluded with clear next steps.
Throughout development, I used Figma Make's AI-assisted coding to build the entire front-end experience. But the AI was executing my UX strategy, not creating it. My role was defining the user journey, specifying interaction patterns based on established usability principles, designing the progressive feedback mechanisms, and architecting the data structure that would serve both AI analysis and sales enablement. The AI accelerated implementation of a human-designed solution.
AI is a tool, not a solution. The real value comes from strategic thinking—recognizing where tools fall short, understanding what's needed to bridge capability gaps, and orchestrating the right combination of technology and design to solve actual business problems. This project demonstrated what I've consistently delivered across 15 years: the ability to identify friction points others overlook, apply domain expertise to solve those problems, and leverage emerging tools strategically rather than superficially.
The Complete System
The Integration Assessment Tool combines intuitive UX design with AI-powered intelligence to deliver personalized recommendations while capturing detailed lead data that transforms sales conversations.
The final experience presents as a professional, branded assessment that guides users through an intuitive progression. Visual app selectors with recognizable logos make tech stack identification fast and accurate—not just saving time, but increasing likelihood that users provide complete, accurate information. The logo-based approach reduces cognitive burden dramatically compared to typing or scanning text lists. The progressive visual feedback system provides subtle but powerful encouragement. As users add apps, small labeled dots appear in the background, creating a visual representation of their growing input. When they map integrations, connecting lines appear between related apps. These elements aren't prominent enough to distract, but they create a positive feedback loop that incentivizes thorough completion. Users aren't just filling out a form—they're building something, and they can see their progress. The recommendation engine analyzes all captured data against Celigo's database of integration patterns and best practices, delivering specific, actionable insights organized by implementation effort and business priority. The branded results page reinforces Celigo's positioning as integration experts with deep domain knowledge, naturally introducing relevant platform capabilities without hard selling. Behind the scenes, the real business value manifests: all data flows to Celigo's CRM as structured fields on the contact record. Sales teams receive unprecedented visibility into each prospect's technical environment, current integration challenges, and stated priorities—comprehensive intelligence that enables accurate lead qualification, smart routing, and substantive first conversations that skip basic discovery and move directly to value discussion.
Design System & BRand Expressions
This project exemplifies strategic AI integration—understanding where tools add genuine value and designing the human-centered systems that unlock it. While this remained a proof-of-concept, it demonstrates the strategic thinking and practical execution that defines my approach to creative leadership: identify the real problem, apply appropriate expertise, and connect technical capability directly to business outcomes.
Strategic Impact
"AI tools are only as valuable as the strategy behind them. The real challenge wasn't building a chatbot—it was recognizing that even sophisticated AI needs the right interface to deliver business value. That's where design thinking transforms technical capability into competitive advantage."
Kendal Richer – Sr. Director of Brand & Creative, Celigo
This project showcases how human-centered design can transform sophisticated AI into a practical business tool: a lead generation engine that captures qualified prospects while enabling smarter sales conversations.
Lead Generation Value: The tool provided a compelling differentiated lead magnet—not generic content, but genuinely personalized recommendations based on each prospect's specific situation. This value exchange significantly increased conversion likelihood compared to standard content offers, while positioning Celigo as integration experts at a critical buyer journey moment.
Lead Quality Enhancement: Beyond basic contact information, the structured assessment captured comprehensive intelligence: complete tech stack details, integration architecture, current challenges, and stated priorities. This enabled sales teams to immediately assess ICP fit, estimate opportunity size, and determine resource requirements—transforming lead qualification from guesswork based on firmographics to data-driven decisions based on technical reality.
Sales Efficiency Gains: Armed with detailed technical context before initial conversations, sales representatives could skip discovery basics and move directly to substantive discussions about solutions, accelerating the sales cycle and improving close rates by focusing energy on truly qualified opportunities.
Brand Positioning Reinforcement: The tool's ability to provide specific, contextual recommendations based on years of customer data demonstrated thought leadership and technical credibility tangibly. Rather than claiming expertise, the experience proved it, strengthening Celigo's competitive positioning in a crowded market.
Strategic Innovation: Building the entire front-end using AI-assisted development tools showcased practical mastery of emerging technologies—not surface-level familiarity, but the ability to evaluate tool capabilities, recognize appropriate applications, and orchestrate AI in service of human-designed strategy. This project illustrated what distinguishes strategic leadership from technical execution: understanding not just how to use tools, but when, why, and in what combination they deliver genuine business value.
The Integration Assessment Tool represents the intersection of AI integration, strategic thinking, and user experience design—demonstrating the ability to identify gaps between technical capability and business impact, then design and build the solutions that bridge those gaps effectively.
Total Time for Design, Development, Testing & Deployment to Staging Environment
