Ascent

• 

  Ascent

• 

  Hero Page

• 

  Features Page

• 

  Meeting Page

• 

  Functioning Chatbot UI and Layout

• 

  Mobile UI

Inspiration

I was inspired by the recent surge in AI adoption across industries. As businesses increasingly look for ways to automate processes and reduce costs, customer support stands out as a key area for innovation. Hiring full-time support staff can be expensive and inefficient, especially for growing e-commerce businesses. I saw an opportunity to build a personalized, AI-powered chatbot that could offer real-time, website-specific support at scale.

What it does

The Customer Service RAG Chatbot is an AI assistant designed for e-commerce businesses. It uses Retrieval-Augmented Generation (RAG) to answer customer queries based on the company's website data and documentation. The chatbot also integrates with tools like Calendly for appointment booking, Stripe for payments, and Supabase for authentication and data storage. It offers a seamless support experience directly on the website, reducing the need for human support.

How we built it

The project was built during a Bolt hackathon using the following tools and technologies:

• Bolt for front-end development and platform integration
• Botpress as the no/low-code chatbot builder
• Supabase for database and authentication
• Calendly for scheduling appointments
• Stripe for handling payments

I also designed a mobile-first, modern UI using Bolt's visual builder, with attention to current UX/UI design trends.

Challenges we ran into

The biggest challenge was integrating multiple platforms and learning each one’s capabilities and limitations. Building an intelligent chatbot with a reliable knowledge base also required significant trial and error. Another hurdle was connecting the chatbot's flow with Stripe, Calendly, and Supabase in a way that felt seamless to the user.

Accomplishments that we're proud of

• Successfully created a fully integrated AI customer support chatbot
• Designed and launched a responsive, mobile-first interface
• Learned how to manage multiple tools within a no-code environment
• Built a scalable solution that has real business value

What we learned

I learned how to use Bolt to build and deploy a complete web application and how to integrate various APIs and platforms. I also gained a better understanding of how retrieval-based AI models work in real-world applications and how to train chatbots using custom knowledge bases. Finally, I saw firsthand how powerful and efficient modern no-code tools have become.

What's next for Ascent

The next step is to productize the chatbot and offer it to small and mid-sized e-commerce businesses. I plan to do cold email automation, instagram dm automation, and look for leads on LinkedIn to do organic marketing. Any funding/prize money I receive will go towards Ad spend for Meta and Google Ads.

Built With

• bolt
• botpress
• calendly-api
• netlify
• stripe
• supabase