Insight+

Inspiration

It all started with a casual chat during the hackathon. We were speaking with Antony, one of the creators participating, and he mentioned how difficult it is to actually understand what their audience feels — especially from comments on video posts or LinkedIn updates. Sure, likes and views are fine, but they don’t tell you how people really feel, or whether you're actually speaking to the right people.

That got us thinking: what if we could turn public comments into something meaningful?

And that’s how Insight+ was born.

What it does

Insight+ helps creators and marketers get the real story behind their posts. Just drop in a public post URL from LinkedIn or Instagram, and Insight+ will:

• Pull all the comments and replies

• Use AI (like GPT-4) to analyze:

  • The sentiment — are people loving it, confused, annoyed?
  • The tone and style — casual, formal, sarcastic, etc.
  • The ICP (Ideal Customer Profile) match — is your content reaching the audience you meant it to?
  • Identify missed engagement opportunities — did someone super relevant comment, and you missed replying?

It then gives you a clean report with:

• 💡 Audience mood summary
• 📊 ICP Match Score
• 🗣️ Language & tone insights
• 🚀 Missed chances + top comments you should act on

How we built it

• Frontend: Next.js to handle URL submissions
• Backend: Laravel for processing and communication with OpenAI
• Scraping: Apify to collect public data from Instagram and LinkedIn
• AI Layer: LLMs (like GPT-3.5-turbo) to handle tone, sentiment, and other nuanced analysis

Challenges we ran into

• Replying to every post is a tedious task for any creator, even a passionate one
• Sarcasm and vague language make sentiment analysis challenging
• Scoring ICP match in a way that feels accurate and insightful
• Figuring out when a creator missed an opportunity to connect
• Making real connections is hard, especially when the audience uses humor or irony in the comment section

Accomplishments that we're proud of

• Built a full end-to-end MVP during the hackathon
• Successfully parsed and analyzed public comments across platforms
• Designed an intuitive ICP scoring system
• Demonstrated missed engagement detection and sentiment reporting
• Created a system that creators can instantly relate to and find value in

What we learned

• Creators aren’t just looking for metrics — they want connections
• You can miss real growth just by missing a single comment from the right person
• LLMs are great at reading between the lines when it comes to emotion, sentiment, and relevance
• Mixing manual inputs (like your ICP) with AI magic creates super-targeted insights

What's next for Insight+

• Analyze multiple posts at once for patterns over time
• Highlight top engaged users and potential influencer collaborations
• Suggest smart replies to comments with GPT
• Build a simple creator dashboard with shareable reports
• Integrate with Slack or CRM tools to turn insights into action
• Include the original post content in LLM analysis for richer sentiment generation

We built Insight+ to help creators go beyond the surface. It’s not just about numbers — it’s about knowing your audience, and making the most of every conversation.

Built With

• laveral
• nextjs
• openai