Angola Explorer AI

Angola Explorer AI turns travel planning into a conversation. Using AI and reliable tourism data, it recommends destinations, attractions, itineraries, and personalized guides for each traveler.

Comment

Inspiration

Planning a trip to a new destination can be a complex process. Angola Explorer AI simplifies this process by combining an intelligent AI agent with a carefully designed tourism database.

Tourism is a growing area in Angola, and as a computer science student, thinking about how technology can be used to boost tourism is something that fascinates me. Beyond that, I also enjoy discovering new places. I imagined how useful it would be for people to travel to unfamiliar locations while trusting a local guide — or even better, first trusting technology to help them make the right decision.

It was exactly from this perspective that I was inspired to develop this project for the Hacktown Build Week.

What it does

Angola Explorer AI connects tourists with local guides. However, the goal is not only to connect people, but to understand that before contacting a guide, the first interaction happens between the user and the AI.

Based on the real data collected during the conversation, the AI suggests a specific guide who matches the user's requirements.

Angola Explorer AI provides an intelligent conversational assistant capable of understanding user intent and generating personalized recommendations.

How we built it

Unlike a traditional chatbot, the AI agent does not generate random information. Instead, it consults the platform's internal knowledge base, which contains:

Tourist destinations Historical information Guide biographies Tourist attractions Photos Reviews Spoken languages Tourism specialties Seasonal recommendations

The retrieved information is then used to generate accurate, reliable, and personalized responses.

Challenges we ran into

Executing this project was very challenging, especially using technologies such as npm, FastAPI, and its dependencies. Although I have experience developing infrastructure with Docker, building a complete MVP alone within a short period of time was a significant challenge.

Accomplishments that we're proud of

I am proud of delivering an MVP within the deadline while preserving the core concept and original vision of the project.

What we learned

I learned a lot about AI agents, RAG, the cost of data required for testing, and the importance of ensuring performance and information reliability.

I also gained a deeper understanding of managing different application layers and using FastAPI resources such as Pydantic schemas to guarantee JSON integrity, FastAPI routers for API organization, Uvicorn for project execution, and SQLAlchemy to translate Python code into SQL queries.

Most importantly, I developed a clearer understanding of the challenges involved in software architecture and AI agent development.

What's next for Angola Explorer AI

More powerful AI Agent with improved RAG capabilities Adding embeddings Multilingual support (PT/EN/FR) GuideAgent A more intelligent RecommendationAgent Conversation memory to maintain context and consistency

Built With

Share this project:

Updates