by iqentity | May 14, 2024 | Development, Tessera
Four months into development, and the extraction results are revealing something I suspected but could not prove until the data confirmed it. Tessera’s real value is not domain expertise. It is meta-heuristics. What Are Meta-Heuristics Meta-heuristics are the...
by iqentity | May 3, 2024 | Development, Tessera
I have been studying other attempts at personal AI systems, and I now understand why they disappoint. The failure is not technical. It is data quality, and specifically, coherence. The Fragmentation Problem Most people who attempt personal AI have data that fragments....
by iqentity | May 2, 2024 | Development, Tessera
I started with off-the-shelf embedding models. They are trained on web-scale data and produce excellent general-purpose vector representations. They also systematically fail on personal knowledge in ways that took me weeks to diagnose. The problem is domain...
by iqentity | Apr 23, 2024 | Development, Tessera
The single most valuable capability in technical remediation is precedent recall. Not “find me a document about Exchange failures” but “walk me through exactly what happened the last time we had a mail flow failure at a financial services client,...
by iqentity | Apr 16, 2024 | Development, Tessera
Raw artifacts are useless to Tessera. An email is a sequence of characters. A PDF is a layout specification. A meeting note is unstructured text. The enrichment pipeline transforms these raw inputs into structured knowledge that the retrieval system can actually work...
by iqentity | Apr 9, 2024 | Development, Tessera
Tessera will never make an API call. Not to OpenAI. Not to Anthropic. Not to any cloud service. The entire system, language model, retrieval infrastructure, knowledge graph, and agentic layer, runs locally on hardware I control. This is not paranoia. This is...
