IntelliGuide: Multimodal AI-Powered Company Assistant

✨ Inspiration

We were inspired by how much time companies waste answering the same repetitive questions—both externally from customers and internally from employees.
We noticed that even with FAQs and chatbots, most systems today are limited to rigid text flows, require developer maintenance, and can’t handle real-world communication like screenshots or voice.
We wanted to fix that.

💡 What it does

IntelliGuide is a plug-and-play multimodal AI assistant that answers company-specific questions using uploaded internal documents.

For employees: It answers HR, IT, and admin questions from internal policy files.
For customers: It provides support for delivery, refunds, and more—based on customer service docs.
It understands text, voice, and images as input and responds clearly in natural language.

All this happens without any code or manual flow setup—just upload your docs, and you're ready.

🛠️ How we built it

Platform: Built with Vible (Lovable’s multimodal AI framework)
Frontend: Created two realistic demo pages
- A mock home shopping site with a customer chatbot widget
- A mock company intranet for internal HR support
Backend logic:
- Integrated document-based QA with natural language understanding
- Enabled support for voice and image input (via Lovable’s multimodal API)
- Developed a separate admin dashboard to manage documents, responses, and chatbot persona

🧗 Challenges we ran into

Designing intuitive multimodal prompts that work reliably across diverse inputs (voice, image, text)
Making sure the chatbot understands document context properly, without hallucinations
Building two separate but consistent demo environments with clear UX boundaries

🏅 Accomplishments that we're proud of

Created a working AI chatbot that supports voice, image, and text across two contexts
Designed realistic demo pages that feel like real business websites—not just prototypes
Built an admin dashboard that non-developers can use to manage all content

📚 What we learned

How to use Lovable's multimodal capabilities to build dynamic AI experiences
The importance of real-world UX when integrating AI assistants into existing systems
How to structure prompts to avoid confusion and maximize document accuracy