Blog Posts

A Deepgram Voice Agent pattern for adding realtime voice to an AI app that already has auth, memory, tools, and transcripts.

What building AI memory taught me about authorization for agents, vector search, delegated access, and why relevance is not permission.

RecallMEM now has a Deepgram voice agent with app-controlled memory. Voice made the memory layer stricter: no quote, no memory.

I thought RecallMEM was a local chatbot project. The model call took almost no time. The real work was persistence, memory, files, tokens, setup, and all the boring systems that make an AI app usable.

The same cycle that made web dev the path into tech is repeating with AI. But the tools are better, the learning curve is different, and the ceiling is way higher.

I miss when coding was hard. The frustration was the craft. Now AI writes the code and developers are having an identity crisis.

How I turned a voice-first AI journal into a Socratic tutor that teaches through conversation, quizzes you until you get it right, and won't let you fake understanding.

Building Speak2Me, a voice-first AI journal that actually knows your story. The hard lessons from building real AI memory systems.

College rankings don't help. I built an AI-powered tool that compares colleges based on actual financial outcomes. Not vibes. Not rankings. Money.

Privacy-first style learning with TiDB vector search. Can you teach an AI how someone writes by storing statistics only?

Most journaling apps are just storage. I built a system that watches my entries, finds patterns in my thinking, and writes my life story in third person - like a biographer following me around.