Zendesk for StartupsUp to 2 years free for startups + AI Agents & Copilot
Promoted
DealFriend — Personal Shopping Companion
AI shopping agent that solves problems, not sells
7 followers
AI shopping agent that solves problems, not sells
7 followers
Alex is your AI shopping friend who searches across Amazon, Walmart, Target, Sephora, and more. Unlike other shopping tools: ✅ Asks about your problem before recommending ✅ Searches all stores (not locked to one) ✅ Has a face and voice (feels like a real friend) ✅ Follows up to see if it helped Perfect for back pain, party planning, home decor, makeup matching, or any shopping problem. Built with Claude API + ElevenLabs voice. Try it: getdealfriend.com
For the makers/developers curious about the tech:
🛠️ Stack:
- Claude Sonnet 4.6
- conversation AI (amazing at understanding nuanced shopping problems)
- ElevenLabs
- voice synthesis (gives Alex her personality)
- Next.js + TypeScript
- Supabase
- auth + outcome tracking
- SerpAPI
- cross-store product search
💡 Hardest challenge:
Getting actual product URLs instead of Google search result pages.
Took 2 days of debugging to map SerpAPI's response fields correctly.
🔧 Next features:
- Virtual try-on (using Fashn.ai)
- React Native mobile app
- Browser extension
Any questions about the build? Happy to dive into details!
Quick question for the community:
What's the worst online purchase you've made because the product description was misleading?
For me, it was that $89 chair that made my back pain WORSE.
That's why I built DealFriend - to ask about YOUR problem before recommending anything. What would you want an AI shopping friend to help you with?
Update: 4 hours in! 🎉 📊 Stats so far:
- 59 visitors from 5 countries (US, India, Pakistan, Sweden, Singapore) - 14% sign-up conversion rate
- Most surprising: Facebook driving MORE traffic than Product Hunt!
The organic word-of-mouth spread is incredible.
People are sharing DealFriend with family/friends across continents. If you've tried it, drop your thoughts below
- what worked? What didn't? Still here answering questions all day! 🛍️
Update (3 hours in):
This is my first time on Product Hunt. Very Exciting...
Climbed from #197 to #103! 🎉
The response has been incredible:
- Reddit post getting great engagement
- LinkedIn: 2,195 impressions organically
- Most surprising:
Facebook driving more traffic than Product Hunt itself The organic word-of-mouth is validating the problem we're solving. For those who've tried it
- what worked? What didn't? Still here answering questions! 🛍️
Behind the scenes moment:
Launched at 2:42 PM (rookie mistake - should have been midnight).
Started at #197. Thought "well, that's that." Then your comments, questions, and feedback started rolling in.
Now at #103 and climbing. This is why I love the builder community.
You give honest feedback and real support.
Thank you 🙏 What else should I share about the build?
Day 2 update! 🎉
Still climbing (#103 → #90) thanks to this community.
Most interesting insight from first 48 hours:
Organic social sharing is outperforming the platform itself.
Facebook + LinkedIn + Reddit > Product Hunt for traffic.
The lesson: Build something people want to share.
The platforms will follow. For everyone who tried it and gave feedback - thank you! 🙏
Especially shoutout to a user who caught the product availability bug. Fixing this week.
What else should I improve?