Insights

Research and AI Lab

The internal research and AI function behind Borta Holdings' investment memos, credit comparisons, route reviews, portfolio monitoring, and decision memory.

Lab mandate

Build tools that improve judgment, not tools that pretend to replace it.

The lab turns documents, market notes, portfolio history, underwriting assumptions, and operating data into organized decision support. Human review remains the control layer. The tools are there to retrieve context, expose contradictions, compare scenarios, and make every major decision easier to audit later.

Research intake

Collects memos, filings, operator notes, market data, logistics updates, and internal observations into a structured research queue.

  • Source and date tracking
  • Asset and decision tags
  • Open-question capture

Memory retrieval

Connects current questions to past decisions, prior assumptions, and similar scenarios so the firm does not relearn the same lesson twice.

  • Decision history
  • Assumption comparison
  • Outcome review

Scenario engine

Frames base, upside, downside, and stress cases for investments, mortgages, routes, and operating-company forecasts.

  • Cash-flow sensitivity
  • Rate and cost shocks
  • Trigger points for action

Tools in development

  • Mortgage term comparator for rate, duration, prepayment, collateral, and stress cases.
  • Fund diligence workspace for fees, manager behavior, strategy drift, and exposure overlap.
  • Shipping route scorecard for ports, timing, fuel, weather, insurance, and geopolitical chokepoints.
  • Portfolio memory map that links every decision to its original assumptions and follow-up notes.

Operating rules

  • AI output must cite the context it used or be treated as a draft, not an answer.
  • Models may organize and compare information, but they do not approve capital.
  • Every tool must make the decision record clearer than it was before.
  • If context is missing, the system should ask for retrieval or review rather than invent certainty.