Trove AI Review: Verdict for PE Teams (2026)
Private equity teams use Trove AI (formerly Mako) to review data rooms and draft portfolio reports, with adoption up 86% among live customers in one quarter.
Trove AI replaces the manual grind of private equity data room reviews and portfolio reporting. Associates at its live customers already run more than 12 tasks a week in the tool, and it fits deal teams and portfolio operations groups that want an always-on research teammate, not another dashboard to babysit.
Trove AI, founded in 2023, is an AI agent platform built for private equity deal teams. It connects to a firm's files, emails and CRM data to independently analyze full data rooms, extract metrics from past deals, and draft portfolio reports in the firm's own format, using a proprietary finance-specific knowledge graph.
Maker: Trove AI · Autonomy: semi autonomous · Maturity: GA
Underlying models: Claude, ChatGPT, Custom (proprietary knowledge graph)
About Trove AI
Trove AI, known as Mako AI until October 2025, is an AI agent platform built only for private equity deal teams. Founded in 2023 by CEO Danny Goldman, a former Bain and Company private equity associate, and CTO Shivaal Roy, a founding engineer on Glean's enterprise search team, the company raised a $7.1 million seed round led by Menlo Ventures with continued backing from Khosla Ventures. Unlike general-purpose AI assistants, Trove is built around a proprietary knowledge graph that models the relationships between deals, portfolio companies, people and metrics the way a private equity analyst actually thinks about them. Trove connects directly to a firm's existing systems, files, emails and CRM entries, and searches thousands of documents without requiring anyone to upload them first. Its agents can independently run full data room reviews, extract metrics buried in past deal files, and draft quarterly portfolio reports written in a firm's own template and tone, tasks that otherwise consume an analyst's week. The upcoming Trove MCP connector will let the same firm context flow into Claude or ChatGPT directly, so associates can query it from tools they already use instead of learning a new interface. The whole system runs cloud-prem, meaning it lives inside the customer's own private cloud rather than a shared multi-tenant service, and is SOC 2 Type II certified. Deal teams use Trove to screen new targets, pull comparable metrics from prior transactions, and prep for investment committee meetings without digging through old data rooms by hand. Portfolio operations teams lean on it for recurring reporting, since it reuses a firm's past formatting instead of starting from a blank page each quarter. It is built specifically for private equity associates, analysts and partners, and has no real use case outside that vertical. Trove does not publish pricing. Firms buy access through a negotiated enterprise contract, typically after a pilot deployment inside their own cloud environment, and the exact cost depends on seat count and the volume of connected data sources. The October 2025 rename from Mako to Trove AI came alongside the seed round announcement. With a small team based in San Francisco, Trove is still an early-stage company, and its near-term roadmap centers on shipping the Trove MCP connector and expanding past its current base of private equity customers.
Pricing
No public pricing. Trove sells through negotiated enterprise contracts priced per seat and per connected data source, usually starting with a pilot run inside the customer's own private cloud before terms are finalized.
Key Features
- Firm-Specific Knowledge Graph: Automatically builds a knowledge graph connecting deals, companies, metrics and people across a firm's internal systems, giving agents PE-specific context standard search tools miss.
- Full Data Room Analysis: Analyzes an entire data room in one pass to surface buried metrics and flag diligence gaps instead of manual page-by-page review.
- Firm-Styled Report Generation: Drafts portfolio reports and IC memos that match a firm's existing template and tone rather than generic output.
- Cloud-Prem Deployment: Runs entirely inside a customer's private cloud and is SOC 2 Type II certified, keeping deal data from leaving the firm's own environment.
- Trove MCP Connector: Plugs Claude and ChatGPT directly into a firm's connected data through the Model Context Protocol, so associates can query Trove from tools they already use.
- No-Upload File Search: Searches files, emails and CRM entries instantly across a connected firm's systems without requiring anyone to upload documents first.
Strengths
- Live customers report associates running the tool more than a dozen times a week, with usage up 86% over the prior three months, per Menlo Ventures' investment writeup.
- SOC 2 Type II certified cloud-prem deployment keeps deal data inside the firm's own private cloud instead of a shared multi-tenant service.
- Backed by nearly $9 million in total funding from Menlo Ventures and Khosla Ventures, giving it more runway than most single-purpose PE research tools.
Weaknesses
- Trove does not publish pricing anywhere on its site, so firms have to go through a sales conversation just to get a quote.
- The Trove MCP connector for Claude and ChatGPT was still listed as coming soon as of the October 2025 funding announcement, so native chatbot interoperability isn't fully live yet.
- It only serves private equity; unlike general-purpose research agents such as Hebbia, it has no application outside deal teams and portfolio operations.
Frequently Asked Questions
How much does Trove AI cost in 2026?
Trove AI has not published a price list. Firms typically start with a pilot run inside their own private cloud, then negotiate an enterprise contract once Trove proves out on a live data room or reporting cycle.
Is Trove AI fully autonomous?
Trove can independently run a full data room review, pull metrics from past deals and draft a portfolio report without step-by-step instructions, but associates still review and sign off on anything that goes into an IC memo or external report.
What are the best alternatives to Trove AI?
Hebbia handles broader multi-step research across finance and legal documents if a firm wants something beyond private equity. Rogo is built for investment banks and covers CIM generation and financial modeling rather than PE portfolio work. Eilla AI targets venture capital and M&A workflows specifically.
How does Trove AI compare to Rogo in 2026?
Rogo's Felix agent is built for investment banking workflows like CIM generation and financial modeling across 250+ firms, while Trove focuses narrowly on private equity data room review and portfolio reporting with a cloud-prem deployment model. Choose Trove if your team is buy-side PE rather than sell-side banking.
How do you get started with Trove AI?
There is no public signup; a firm requests a demo through Trove's website, which typically leads to a pilot deployment inside the firm's own private cloud before contract terms and pricing are finalized.
Top Alternatives
- Rogo: Pick Rogo if your team runs investment-banking workflows like CIM generation; pick Trove if your team does private equity deal and portfolio work.