Xplainr

Inspiration

Most small businesses want what large enterprises already have — an intelligent, 24/7 customer service agent that actually understands their products and policies. But creating one usually requires technical knowledge, API integrations, and expensive infrastructure. We wanted to make that process as easy as uploading a website link. That’s why we built Xplainr, an app that lets anyone instantly deploy a custom AI agent trained on their live website and company documents, no coding required.

What it does

Xplainr allows users to instantly create an AI customer service agent that understands their business. The user simply enters their company name and website URL. Xplainr automatically crawls the website, extracts relevant information, and builds a knowledge base around it. Users can also upload documents such as FAQs, policies, and manuals to provide deeper context. Within seconds, the platform configures an AI agent capable of answering customer questions with accurate, real-time information. Whether it’s a small business or a large enterprise, Xplainr makes deploying a company-aware AI support assistant as simple as clicking “create.”

How we built it

The backend was built using FastAPI, which handles company onboarding, web crawling, document processing, and dynamic AI configuration. On the AI side, Xplainr uses ElevenLabs’ Conversational AI Agents Platform to power real-time chat and voice interactions, Gemini API for generating embeddings and performing retrieval-augmented reasoning, and ChromaDB for storing and searching embeddings derived from websites and uploaded files. Each time a new company is added, Xplainr automatically updates the agent’s system prompt, dynamic variables, and knowledge base, giving every business a unique and context-aware AI assistant.

Challenges we ran into

One major challenge was keeping multiple APIs synchronized while dynamically updating live agents. Crawling websites efficiently while maintaining high-quality retrieval performance required careful optimization. We also faced SDK versioning issues, particularly with Python 3.13, which introduced circular import errors. Additionally, handling large and unstructured company data required designing a robust pipeline for extracting, cleaning, and embedding information consistently.

Accomplishments that we're proud of

We built an end-to-end pipeline that can transform any website into a fully functional AI agent in under a minute. We achieved true one-click deployment with no manual configuration required. The system scales easily, working equally well for small local businesses and large enterprises. We also successfully integrated web crawling, document retrieval, and conversational AI into a single seamless platform that feels effortless to use.

What we learned

We learned how to leverage ElevenLabs’ Conversational AI API to create and update agents programmatically and how retrieval-augmented grounding ensures that AI responses stay factual and relevant to a company’s live data. We gained valuable insight into designing AI systems that hide technical complexity behind a simple user experience. And perhaps most importantly, we learned how to debug and adapt rapidly while working with evolving SDKs and APIs under tight deadlines.

What's next for Xplainr

Next, we plan to take Xplainr beyond the hackathon stage and develop it into a scalable startup. Our immediate goal is to refine the product into a fully self-service platform where businesses can launch and manage their own AI agents in minutes. We are also beginning preparations to apply to Y Combinator, as we believe Xplainr has the potential to redefine how businesses—especially small and medium-sized ones—build and deploy customer service automation. Through Y Combinator, we aim to gain mentorship, funding, and exposure to accelerate our development, grow our user base, and form key partnerships with businesses seeking reliable, low-cost AI infrastructure. Our long-term vision is to make Xplainr the standard for instantly deployable, intelligent customer service agents across industries worldwide.