LightfieldAI-native CRM that builds itself and does work for you
Promoted
Reviewreply ai
"Reply to Google Reviews in 1 Click with AI"
3 followers
"Reply to Google Reviews in 1 Click with AI"
3 followers
"Free AI tool that helps small business owners reply to Google reviews instantly. Supports English, Arabic, Urdu & Hindi."
Cool idea, the multi-language auto-detect is smart. I'm building something similar for website chatbots (AI that answers from a knowledge base) and the tone-matching piece is tricky to get right. How are you handling cases where a negative review is actually constructive feedback vs just someone venting? Feels like those need very different reply styles.
@cuygun Thanks so much for the kind words! 🙏
Great question! Right now we detect sentiment using keyword analysis — words like "slow", "rude", "terrible" flag it as negative, while "great", "amazing", "loved" mark it as positive.
You're absolutely right that constructive feedback vs venting needs different handling. That's actually on our roadmap — we plan to use more context-aware AI analysis to distinguish between:
Someone genuinely upset (needs empathy first)
Someone giving constructive feedback (needs acknowledgment + action)
Would love to hear how you're handling it in your chatbot! 🚀
@abdul_karim40, Yeah keyword-based sentiment is a solid starting point, gets you 80% of the way there. When you move to context-aware analysis you'll probably see a big jump in quality, especially for those reviews that use positive words sarcastically ("great, loved waiting 45 minutes").
On our side we use a system prompt that tells the AI to match the user's tone and language, but we also gate it so the chatbot stays within the knowledge base and doesn't make promises the business can't keep. That last part turned out to be harder than the tone matching honestly, users will try to push the bot into saying things like "we'll give you a refund" when that's not in the KB.
Your multi-language angle is interesting though, we support multiple languages too and the tone nuances across languages are a whole different challenge. Good luck with the roadmap!
@cuygun This is gold, thank you! 🙏
The sarcasm point is spot on — "loved waiting 45 minutes" would completely fool keyword analysis. That's going on the priority list!
The knowledge-base gating challenge you mentioned is really interesting — keeping AI from making promises the business can't keep is something we'll face too when we add business profiles.
Really appreciate you sharing your experience. Let's stay in touch — would love to see what you're building! 🚀
@cuygun can we chat on insta @ malik_mubeen_2009 , this is my you can text me here for future sport or working together
@abdul_karim40, appreciate the interest! If the sentiment analysis solves a real pain point for businesses you've talked to then yeah, worth building. The key thing at this stage is getting it in front of real users and seeing if they stick, don't over-build before you have that signal. Happy to keep exchanging ideas here on PH, best of luck with it!