Founder Engineering
Building Highlyt: A PDF Reader for the LLM Era
A case study on building Highlyt, a PDF reader and typed knowledge graph that connects reading workflows with Claude and ChatGPT through MCP.
Problem
Most reading tools stop at capture. Highlights become static notes, and the reasoning behind a source gets separated from the document. Highlyt was built around a different premise: readers need a way to preserve relationships between ideas, not just collect passages.
System
The product combines semantic PDF highlighting, typed links between highlights, cross-document clusters, and MCP access so Claude and ChatGPT can read from and write to a user's library. The technical challenge is keeping annotation, graph, and LLM workflows consistent across web, extension, mobile, and MCP clients.
Original Write-up
I wrote the longer product essay around the core belief behind Highlyt: I do not want an AI that reads for me, I want one that reads with me.