What AI Sees in Your Portfolio That Your Spreadsheet Doesn't
How AI-assisted analysis surfaces regime-conditional risk, hidden factor exposures, and portfolio construction options that traditional tools miss.
AI × Asset Management
Orikai writes about what it actually takes to make AI work inside an asset management firm — the workflow, the scepticism, the data problems, and the parts vendors don't tell you about.
Subscribe on Substack →Writing
Published on Substack. Focused on solutions, not tools.
How AI-assisted analysis surfaces regime-conditional risk, hidden factor exposures, and portfolio construction options that traditional tools miss.
And I built a working version of it using Keppel Data Centres to see how far it could go.
And what the experiment taught me about where AI moats in finance actually sit.
A framework for how buyside firms in Southeast Asia actually use AI.
Solutions
Each tool exists to answer a specific question a fund manager actually has.
Processes earnings call transcripts and extracts sentiment, management tone, guidance flags, and theme shifts — so a small team can cover reporting season without reading every transcript in full.
Open tool →A RAG system built over Keppel DC REIT's filing corpus. Ask questions about the company and get answers grounded in actual documents — not model hallucinations. A proof of concept for company-specific RAG in SEA.
Open tool →Rolling correlations, regime detection, factor exposure, and tail risk for SGX-listed portfolios — with plain-English interpretation. Built for PMs who need risk framing, not just risk numbers.
Open tool →Screens Singapore REITs across key metrics without requiring Bloomberg access. Built for boutique and family office investors covering the SGX market with lean data infrastructure.
Open tool →Aggregates market intelligence from multiple sources into a structured daily briefing — formatted for a CIO's inbox. Designed around what a senior investor actually needs to see first thing.
In developmentIngests broker PDFs across banks and dates. Tracks consensus drift, target price revisions, and rating changes over time — with alerts when the street diverges from a stored house view.
In developmentField Notes
Things I've noticed that don't fit into a full article. Updated as I go.
About
Orikai is a writing and practice focused on AI implementation inside asset management firms — specifically the parts that are harder than the pitch decks suggest.
I'm a corporate development professional with 8+ years of experience in Southeast Asia, now working inside an asset management firm to understand what it actually takes to deploy AI in this context. I'm not an ML researcher. I'm someone who understands how investment teams work and has learned enough to build things that fit that world.
The honest version of AI implementation is mostly a people and process problem. Orikai is where I write about that honestly — for fund managers, CIOs, and operators who are trying to figure out where to start.