Short, dated notes on real investors and firms using AI, from X, Reddit and the news. The best ones graduate into full use cases.
After a four-month proof of concept with AWS, Korea's National Pension Service says AI agents now handle central-bank policy analysis that took a portfolio manager about eight hours in roughly 10 minutes, and automate much of the earnings-commentary work several PMs spent about 180 hours on each quarter. The PoC used only public data; NPS plans to widen AI use across fund management research.
Speakers at the Circulus Family Summit in New York said family offices are starting to feed governance and mission documents into their AI systems, and called an ontology layer that maps the office's real workflows the biggest win. Eton Solutions' CEO warned against vibe-coding core systems like a CRM in-house, citing edge cases and security.
Mosaic Ventures, a partner-led VC firm with no associates or analysts, re-scores tens of thousands of startups every two weeks using external data and its own CRM signals, weighted separately for each partner, then runs an AI screening pass before the top-of-funnel meeting. The system was the main catalyst for about 15-20% of the firm's first meetings over the past year; the post also covers what hasn't worked, like semantic thesis matching.
Jas Khaira, global head of Blackstone N1, explains how the firm works with portfolio company leaders to rethink workflows and build AI into operations.
The 2026 RBC and Campden Wealth report finds half of North American family offices now see AI as essential to workflows, with cybersecurity the top operational concern.
Artmarket.com's H1 2026 report positions its Intuitive Artmarket and Blind Spot AI systems as the core of an 'AI-first' Artprice through 2030.
The world's third-largest pension fund plans to deploy AI to monitor ESG issues across the companies it holds, starting next year.
An advisor who spent a day with a private equity real estate firm describes partners producing decks, models and LP updates with AI instead of an analyst class.
The angel group's series on its own AI systems: specialist agents handle research on hundreds of applications a year, and members add judgment at defined checkpoints.
Jenny Johnson said the firm's agents work across its investment teams today, and that the next step is orchestrating many agents together. Human judgment stays accountable for decisions and unusual market events.
NPS and LG AI Research are building and testing an AI currency-analysis service on macro indicators, market data, news and internal fund data, with plans to widen AI use across fund management.
Bloomberg reports investors are pressing Apollo, Blackstone and peers for clearer disclosure of AI exposure and risk across their funds.
Institutional Investor's look at how big pension plans with record asset growth are adding AI and data talent for research, risk monitoring and operations.
The software-focused buyout firm reports running about 1,600 AI projects across its portfolio with $260 million of budgeted EBITDA impact, one of the clearest numbers yet on PE value creation from AI.
The AI-native fund administrator's CEO says a new institutional client switched after reference calls with two existing customers. Agents process emails and entries; CPAs review the output.
The longtime hedge fund manager built Bracket22 around specialist agents (one for technicals, one for quant strategies, one coordinating) while he keeps the final call on every trade. He told CNBC labor-related costs fell from about $5 million a year to roughly $30,000 to $40,000.
Aleta pulls together 2026 survey data showing a gap between family offices' AI investments and their own operational use.
On the Dry Powder podcast, Stavros describes KKR's diagnostic grid for portfolio AI, embedding AI talent with deal teams, and where vendor trials have and haven't paid off.
Executives at an industry event described using AI in portfolio operations; Blackstone cited AI deployment across 1,400 portfolio companies and a fivefold speed-up in lease processing.
The California hedge fund's co-founder says point solutions fragmented and in-house tools became a full-time job, so the fund moved to a hosted agent platform customized to its process.
Business Insider on how Wall Street teams use Hebbia's updated Matrix to work through filings and research at scale.
AngelList shares how fund managers use its read-only connector to ask Claude and other assistants live questions about fund and LP positions.
After a stalled CIM-analysis pilot, the 150-person firm rebuilt its rollout around custom GPTs for CIM write-ups, board decks, NDA review and financial summaries, wired into Outlook, SharePoint and DealCloud. Per the implementation partner, CIM summaries went from 4+ hours to under two minutes.
A Lerer Hippeau investor's comment, quoted in the Standard Metrics raise thread: structured portfolio data makes AI analysis of portfolio company KPIs straightforward.
Every item links to its source. Items marked "Pending review" are awaiting editorial sign-off.
Artprice
Aleta
Hebbia
AngelList
ChatGPT
Standard Metrics