Just-in-time ontology service: entity-resolution and data-transformation pipelines built on demand, priced per decision
Instead of forcing all data into one universal schema ahead of time, build the transformation only when the question is known — then preserve the ontology with its receipts (evidence, definitions, how far each answer was checked) so it can be reused for the next job. Sold as a service that resolves messy records (e.g. deciding whether campaign finance donors, companies or committees are the same entity) using cheap decision models, with self-optimizing agent loops tuning cost, speed and accuracy.
online BOTH AI Agents & Automation
From AI Engineer — Just-in-Time Ontologies: What Cheap Intelligence Means for Data — Hrishi Olickel, Southbridge AI at 01:14
Problem: Universal schemas and ontologies fail because there are four schemas (source, storage, decision, and the model's prior) that are never known at the same time except at the moment of decision — a 'customer' means something different to billing, support and compliance. Agentic transformations produce confident wrong answers (a net $784,000 became $5.3 million of debt, signed away by multiple agreeing models).
For: Organizations with messy multi-source records that need reliable answers for specific decisions — compliance, billing, research — where a number must be trustworthy rather than merely produced; governments, regulated firms, data-heavy businesses.
Products from this video
Examples
- Ohio campaign finance entity resolution: all 23,317 statewide donor cases labelled with evidence and check depth; originally projected at about $10,000, delivered with final ground truth 10 days after the new model launched for $7.95 of an $800 pre-set budget, with checking costing roughly four times as much as doing ($545 on second opinions vs $128 on first answers)
Other takes on AI agent platforms and compute access
- Autonomous AI agents that perform administrative labor for underserved small businesses, starting with dental practices
- A personal, open-source AI agent that can be operated through messaging channels such as WhatsApp and Discord
- AI virtual staging service for real estate listings
- Enterprise AI concierge for customer, sales, and operational workflows
- AI appointment-scheduling assistant for healthcare providers
- Autonomous SEC-filing research desk with self-graded scorecards