I built a simple, totally free family tree app after struggling to map my own family.
I know Ancestry, MyHeritage, FamilySearch, etc. already exist. The issue for me was not historical records or DNA research. It was just creating a clear, shareable tree for a very large real-world family.
My family tree is already 300+ people across 5 generations, with lots of branches and relatives who know different pieces of the story. Most tools I tried either felt too research-heavy, too complicated, or did not scale well for the way my family is structured.
So I built FamilyTreeIQ to be simple, free, focused on building and sharing the tree itself, usable for large families, and easy enough for relatives to understand.
I’m not trying to replace Ancestry or serious genealogy tools. This is more for families who want a clean, shared way to map “who’s connected to who.”
I’d love blunt feedback, especially from anyone with a large or complicated family tree:
Does this feel useful? Is the idea clear? What would make you trust a free family-tree app with your family info?
I got sick of getting spam calls from the same company 4+ times a day for almost two months straight. They kept ignoring the Do Not Call registry, even though they claim to have it implemented.
So I decided to build something to fight back: an AI that takes over and wastes their time instead.
Watch it in action here: https://www.youtube.com/watch?v=AldNjRm4gzQ
I put it together using a mix of Twilio, OpenAI, ElevenLabs, Deepgram, plus web sockets, audio compression, and VOIP. It's been a fun project to work on.
Right now, I’m not ready to make it public (because it does have some costs to run), but if enough people are interested.
Let me know what you think!
20+ year senior software engineer here, sharing my experience with LLM/AI agents.
TL;DR: LLMs are genuinely useful and many experienced developers use them daily, but they’re being oversold. They work best as tools that amplify expert developers, not as replacements for them. They can generate and refactor code well, but they hit a hard ceiling on deep reasoning, architecture, and optimization. In expert hands, they save time; treated as autonomous engineers, they produce confident-looking but low-quality results.