About Services NVS Use Cases Blog Email us

Practitioner notes on AI for financial services.

Notes from twenty years on the buy side, now spent building AI for it.

October 19, 2026

How AI Strengthens Technical Due Diligence in Private Equity

AI lets a deal team cover far more of a target's code, data room and engineering history within a short diligence window. It works best when pointed at the specific questions the investment committee needs answered, with experienced reviewers signing off on every finding.

October 12, 2026

Research Copilots a Portfolio Manager Can Actually Trust

A research copilot earns a portfolio manager's trust when every claim traces to a source, it admits what it cannot find, and it has been tested against the firm's own research history. Here is what that requires in practice.

October 5, 2026

Taking an Agentic Research System from Prototype to Production

An agentic research system is ready for production when its outputs trace back to sources, its quality is measured against a fixed evaluation set, its permissions are scoped, and a named person owns every decision it informs. This post covers what that takes in practice.

September 28, 2026

What an AI Readiness Assessment Finds at a Hedge Fund

An AI readiness assessment at a hedge fund usually finds that data history, undocumented workflows, governance, and ownership limit AI more than the models do. This post covers what the review examines, what it tends to find, and what the final output should contain.

September 21, 2026

What Production AI on a Fund's Stack Actually Looks Like

Production AI at a fund is a small set of narrow, monitored systems embedded in existing workflows, each with a named owner and a measurable output. Here is what that actually looks like once you get past the pilot.

Where AI Creates Real Value Across a Private Equity Portfolio

AI pays reliably in three places across a PE portfolio: fund-level diligence and monitoring, repeatable functions common to every holding, and a small number of revenue workflows in the largest companies. A practitioner's view on what works, what does not, and how to sequence it across the hold period.

Automating Regulatory and Client Reporting for Asset Managers

Reporting automation succeeds when you fix the data layer and the reconciliation logic first, then generate documents from it. Here is the architecture, the controls, and the sequencing that hold up under regulatory and client scrutiny.

AI for Technical Due Diligence in Private Equity

AI has become genuinely useful in technical due diligence, mostly for reading code and documents at a scale humans cannot match inside a three-week window. Here is what it does well, where it misleads, and how to run it on a live deal.

How To Know If An AI Project Is Worth Funding In Financial Services

How To Know If An AI Project Is Worth Funding In Financial Services

Research Copilots a Portfolio Manager Can Actually Trust

A research copilot earns trust when every claim it makes is traceable to a source, scoped to your data, and reproducible. Here is how to build one that survives contact with an investment committee.

Taking an Agentic Research System From Prototype to Production

A prototype that answers your questions in a demo is not a production research system. Here is what closes the gap: deterministic orchestration, source-grounded outputs, evaluation harnesses, cost controls, and human review gates.

What an AI Readiness Assessment Finds at a Hedge Fund

A readiness assessment at a hedge fund almost always finds the same five gaps: messy data lineage, unclear ownership, no evaluation method, security exposure, and a backlog of low-value pilots. Here is what each looks like and how to fix it.

AI Research Assistant For Hedge Funds: Turning Scattered Research Into Decision-Ready Work

AI research assistants help hedge funds turn scattered filings, transcripts, memos, and models into decision-ready research. Built with governance from day one, they improve retrieval, thesis maintenance, and investment committee prep without replacing analyst judgment.

What Production AI on a Fund's Stack Really Looks Like

Production AI at a buy-side firm is plumbing that lives next to the OMS, the data warehouse, and the compliance log. It earns its place by being narrow, observable, and reversible.

How Hedge Funds Use AI for Risk Management

Hedge funds use AI to detect regime shifts earlier, capture nonlinear exposures, generate forward-looking stress scenarios, and read unstructured data for tail risk. Here is how the buy side puts each to work, and where model discipline matters.

The AI advice changes when the advisor has run the book

Most AI consultants met financial markets on their first engagement. Here is what changes when the person advising you has actually managed a portfolio.

See where AI actually pays off in your firm.

A focused 30-minute session on your highest-value AI opportunities.

jg@novarize.ai