Axel the AI Chatbot

Freelance work conducted for a business owner. More information about this project and the development process can be shared verbally.

Introduction & Objective

My role: Understand project mission

Small business owners are constantly stretched thin. They juggle customer questions, inventory, and day‑to‑day operations - often without the resources to hire additional staff. Many of the questions customers ask (“Where are you located?”, “What time do you close?”, “Is this in stock?”) are simple and repetitive, yet they consume valuable time.

My objective was to help Industrial Bicycles reclaim time, reduce repetitive support tasks, and improve customer confidence by creating a custom AI chatbot, named Axel, that provides real‑time, 24/7 assistance without adding payroll or operational overhead.

Research for User Needs

My role: User Research Lead

An extensive benchmarking report and comparative analysis was conducted. I identified key user pain points:

Customers struggled to find product information.
Many asked the same questions repeatedly.
The business owner lacked time to respond quickly.
Customers wanted fast, confident answers before making a purchase.

I translated these insights into clear need statements:

Users need quick access to accurate information.
Users need help comparing products.
The business needs to reduce repetitive inquiries.
The business needs a scalable, low‑maintenance support solution.

Research for AI Platform

My role: Lead Developer & Researcher

To build Axel, I researched:

AI agent capabilities
Prompt engineering best practices
How to structure knowledge sources
How to embed scripts into existing website HTML
How to integrate with tools like SharePoint, Miro, and PowerBI for real‑time updates

This research guided the technical foundation for the chatbot.

Ideation & Concept Design

My role: Lead designer

Concept sketches included:

A conversational UI embedded on product pages
A knowledge‑driven agent trained on product specs, store info, and FAQs
A tone that felt friendly, helpful, and aligned with the brand

I also designed the system architecture:

Knowledge sources
Prompt structure
Guardrails
Response formatting
Error handling

Prototypes & Evaluation

My role: Lead Prototyper & Lead Tester

I built early prototypes of Axel using iterative testing:

Initial versions struggled with accuracy and link handling
I refined prompts, improved instructions, and cleaned up product data
I tested the chatbot with real customer questions
I embedded the script into the site’s HTML through trial, error, and research

Evaluation focused on:

Accuracy of responses
Tone and clarity
Ability to guide users to the right product
Reduction in repetitive questions
Ease of use for non‑technical customers

Each iteration improved reliability and user experience.

Final Product

My role: Product Owner & Developer

The final version of Axel, the Industrial Bicycles AI chatbot, now:

Provides real‑time, 24/7 customer support
Answers common questions instantly
Helps customers compare products
Reduces repetitive tasks for the business owner
Integrates with PowerBI, Miro, SharePoint, and an AI agent backend for real‑time updates
Is fully embedded and deployed on the website

Axel is now a reliable digital team member, and genuinely in the running for “employee of the month.”

Impact & Reflection

This project demonstrated how accessible and transformative AI can be for small businesses.

Impact:

Reduced repetitive customer inquiries
Improved customer confidence and purchase readiness
Saved the business owner time without increasing payroll

Reflection:

Understanding users is just as important as prompt engineering
You don’t need to be a programmer to build powerful AI tools. Curiosity and persistence go a long way
Small details (tone, links, instructions) dramatically affect user experience
AI is a game‑changer for lean teams when implemented thoughtfully