
A LinkedIn ghostwriter fueled by your data and the news
Most LinkedIn tools are open-loop: you provide a prompt, and it spits out generic AI slop that requires heavy editing. Isograph closes this loop by merging your internal performance data with external industry news. It ingests your LinkedIn analytics to learn your unique voice while a Scout agent monitors real-time trends to solve the cold-start problem of what to write today. The result is a stateful agentic system that evolves its strategy and drafts based on what actually converts.

Screenshot 1

Screenshot 2

Screenshot 3
Sign in to your GMA account to leave a review and help others.
GMA Community
"Hi Product Hunt! 👋 I am Sukhendra, and I built Isograph because LLMs weren't producing right content for me for Linkedin. We have all seen the generic, hollow content most AI tools spit out. It usually takes more time to edit the "AI-ness" out of a post than it would have taken to write it from scratch. The root cause? Most tools are stateless, they don't know who you are, what you’ve said before, or what actually resonates with your specific audience. Isograph is my attempt to close that loop. I wanted to build a "System of Intelligence" that feels like a high-end agency, not a text box. It uses a multi-agent architecture to handle the heavy lifting: 1) The Scout: Monitors real-time industry news so you never have a "cold start" morning. 2) The Reflection Agent: Ingests your actual LinkedIn analytics to learn your "Voice" and what converts. 3) The Bridge & Guardian: Ensures every post strategically connects your brand to the news while staying within your safety constraints. I have just moved this from a prototype to a live production version at http://isograph.vercel.app. You can upload your own analytics history right now to see how the "Reflection" layer changes the output quality. I’d love to hear your thoughts on the "Closed-Loop" approach or any feedback on the drafts the agents generate for you! Happy Linkedin post generation, Sukhendra"