ai-receptionist

• 

  BW

Inspiration

BlvckWall AI Receptionist was inspired by seeing how much time and opportunity businesses lose just because nobody’s available to answer the phone or follow up with clients. I wanted to build something that goes way beyond basic call routing: a real AI assistant that can handle the heavy lifting, let owners customize how calls are handled, and tie everything together in one smooth dashboard.

What it does

AI Receptionist lets business owners create AI-powered phone agents that handle both inbound and outbound calls. It manages multiple business numbers, schedules appointments, transcribes calls, translates conversations into different languages, generates video summaries, and even audits all activity for compliance. The dashboard brings all of this together with analytics, agent management, HR tools, and real-time error tracking so owners never miss a beat.

How I built it

I built the dashboard using Next.js, React, and TypeScript for both the frontend and backend. ElevenLabs was the main engine for the voice agents, powering both incoming and outgoing calls. I also used Retell AI for some call handling during development, but the core voice features rely on ElevenLabs. For advanced features, I integrated APIs like Tavus (for video summaries), Lingo (for translation), Algorand (for call auditing), and Pica OS (for business card creation and verification). Sentry was added for error tracking, giving owners and their teams live updates when something breaks. Supabase was started for database and authentication, but for the demo, everything runs on local storage to keep things fast and simple.

Challenges I ran into

One of the toughest parts was connecting so many different APIs and sponsor tools into a single dashboard, especially when you want it to feel seamless and not like a bunch of glued-together features. There were also a lot of bugs with Supabase integration, so to meet the hackathon deadline, I kept things in local storage for the demo. Testing and debugging real phone call flows, especially with ElevenLabs’ live audio, meant a lot of trial and error.

Accomplishments that I'm proud of

I’m proud that I managed to build a fully working dashboard that covers real business needs, not just a simple call bot. Owners can spin up AI agents, manage numbers, see analytics, generate videos, and use translation—all in one place. Every sponsor feature is actually integrated and working in the demo, not just a mockup.

What I learned

I learned a ton about managing async API flows, handling voice data, and building UIs that actually help business owners instead of just looking pretty. I got better at choosing when to polish and when to move on, and how to show off what matters most in a tight timeframe.

What's next for ai-receptionist

The next step is to finish the Supabase integration for real accounts and persistent data, clean up bugs, and polish the overall user experience. After that, I want to roll out the product to actual businesses, add more automation features, and eventually scale up with a team to make this a production-ready platform. If funding comes in, I’ll double down on new features and push hard to make AI Receptionist the new standard for business communication.

Built With

• 21st.dev
• algorand
• clsx
• crypto-js
• date-fns
• elevenlabs
• framer-motion
• google-calendar-api
• ipfs
• javascript
• lingo
• local
• local-storage
• lucide-react
• microsoft-outlook-graph-api
• netlify
• next.js
• node.js
• npm
• picaos
• postcss
• postgresql
• radix-ui
• react
• react-hook-form
• recharts
• retell-ai
• row-level-security-(rls)
• sentry
• session-storage
• shadcn/ui
• sql
• supabase
• supabase-auth
• supabase-realtime
• tailwind-css
• tailwind-merge
• tavus
• typescript
• uuid
• vercel
• webpack
• zod