Kepler vs V7 Go

For private equity and credit teams who need diligence and monitoring on the firm’s own definitions, with the public record already read and every number computed in code and reproducible, Kepler is the stronger choice. V7 Go runs document workflows on a Context Graph of the firm’s files: agents that screen CIMs, work through data rooms, complete DDQs and draft memos, with each output cited to its source page. Kepler builds one living model of your firm from every document you hold plus the public record already read, applies your definitions everywhere, and answers from an engine where the model never writes a number.

We build Kepler. Facts about V7 Go come from V7’s website and announcements as of November 2026, linked inline.

The short version

• V7 Go automates document-heavy workflows across private markets, insurance and real estate: a library of agents, a Context Graph of the firm’s files, Slides for decks, and more than 300 integrations and MCP connections.
• Kepler is enterprise intelligence for investment firms. Our engineers build your firm’s ontology, extraction models we train read every page at ingest, a deterministic engine computes each figure with its formula and sources, and agents run on schedules and events.
• The difference is what sits under the answer. V7’s Context Graph links a firm’s entities and relationships to source evidence. Kepler adds the public record already read (26M+ SEC filings with XBRL, 40M+ IR documents and transcripts), your firm’s definitions written into the ontology, statements as filed or as restated, and an engine where code computes every number and the model never writes one.

Kepler vs V7 Go side by side

	Kepler	V7 Go
Built around	Your firm’s ontology, shared by every question, agent and report	A Context Graph of firm files (entities, relationships, source evidence) that feeds agent workflows
Public data	26M+ SEC filings with XBRL, 40M+ IR documents and transcripts, market data, news, already read	Connected sources; PitchBook through its MCP connector
Reading	Every page of every connected document at ingest, plus the public record already read	Firm files extracted into the Context Graph; pricing scales with pages processed
Where each number comes from	Pulled from the source and computed by code; the model never writes one	Extracted by agents, with code for calculations, per V7
What a citation opens	The exact page, line or cell, checked before you see it	The source page, highlighted
Financial statements	Assembled from each filing’s XBRL, with no model in the path	Extracted from the documents provided
As filed vs as restated	Either, on any date	Not described by V7
Your firm’s definitions	Written into the ontology and applied everywhere	Configured per workflow in fields, criteria and templates
Same question twice	The same number, with the same sources	Not stated publicly
When sources disagree	Both shown, with sources, and the conflict flagged	Not stated publicly
Work that runs on its own	Agents on schedules and events, acting as their owner	Workflows that run as documents arrive
Where it works	App, Excel, PowerPoint, Word, Claude, ChatGPT, Slack, email	V7 Go app and Slides, MCP and API
Build on it	API, MCP server, Python SDK, webhooks	MCP, API and 300+ integrations
Built for	Investment firms: hedge funds, PE, credit, asset managers, allocators	Private markets, insurance and real estate

What V7 Go is

V7 Go is V7 Labs’ platform for document-heavy work, aimed at three industries. In private markets it covers deal screening, due diligence, investment memos and portfolio reporting; in insurance, submission ingestion, underwriting triage and claims; in real estate, deal analysis and lease abstraction. Teams start from a library of agents written by industry experts or build their own, and V7 lists more than 300 integrations and MCP connections, including PitchBook’s Premium MCP Connector (August 6, 2026) and an Affinity partnership for deal documents (September 16). V7 Go Slides (April 2026) builds decks in a firm’s own template, with editable charts and a proofreading agent that checks each figure against internal documents or the web.

The center of the platform is now the Context Graph, which V7 describes as firm memory: it connects fund reports, data rooms, memos, emails, entities, relationships and source evidence so that each workflow starts with history attached. In an OpenAI customer story (September 2026), V7 explained that Go carries that context through defined steps, using models for reasoning and code for calculations, with review points and an audit trail, and that its solutions engineers build workflows with each customer as part of implementation. V7 says every output is cited to its source page. There’s no public price list: the annual agreement is built from platform access, the number of decision-making users, and the volume of pages processed.

What Kepler is

A model on its own has two gaps that matter in a deal. It writes numbers as text, so a figure can be wrong and look right. And it doesn’t know the firm: your prior deals, your portfolio, your definition of net revenue retention, the memo format your IC expects.

Your firm’s ontology. Our forward-deployed engineers built this layer before at Palantir. They connect your data rooms, SharePoint, OneDrive, Box, Egnyte, Outlook, Slack, CRM, research notes and models read-only and build one living model of your firm: every company, fund, deal, sponsor, lender, manager and document resolves to a single record, each relationship cites the sentence behind it, and your house terms carry one definition. It knows time, so a figure is available as reported, as restated or as of any date. Kepler also ships with an ontology of public companies built from SEC filings.

Every document read in full. Extraction models we train read every page at ingest, scans, tables and exhibits included, and store each fact with its value, scale and unit beside the passage it came from. On 40 held-out filings, our readers got the scale of 99.4% of tagged facts right, as Stop Paying to Forget reports. The public record is already there: 26M+ SEC filings across 20+ form types and 40M+ IR documents and earnings call transcripts for 14,000+ companies, so a target’s listed competitors sit next to its data room.

A deterministic engine. Code pulls and computes every number. Financial statements assemble from each filing’s own XBRL with no model in the path, every calculation runs over cited inputs with its formula shown, and every citation is checked against its source before you see it. Ask the same question twice and you get the same figure from the same lines.

Agents and playbooks on the right models. Agents run on schedules and events (a new upload to the room, a Q&A response, a monthly package, a filing from a peer), acting exactly as their owner and never taking instructions from what they read. The IC memo in house style, the screening checklist and the quarterly review are saved once and applied every time. We route each task to the best model for it, small models we train for reading and frontier models for the question and the write-up, and swap in better ones as they ship. All of it reaches the Kepler app, Excel, PowerPoint, Word, Claude, ChatGPT, Slack and email, and your engineers can build on our API.

Not the right fit if: your team needs one extraction workflow for insurance submissions or lease abstraction, outside investment work.

“Every other AI tool I’ve tried misses what’s buried in the footnotes. Kepler has read every filing, so nothing slips past it.” (Managing Director, multi-billion dollar fund)

The same data room, two ways

A deal team gets access to a 1,400-file data room on a Monday and asks: “For Project Larch, compare customer concentration, gross margin and net revenue retention with the three vertical software deals we closed since 2022, on our definitions. List every customer contract with a change-of-control clause, and draft the screening memo.”

In V7 Go, a diligence agent works through the room, extracting financials and contract terms with a citation to each source page, and the Context Graph brings the firm’s prior deals and memos into the run. Slides or the memo agent draft the output in the firm’s template. In a trial, check where the prior deals’ figures come from (a CIM, a final model or post-close reporting), which definition of net revenue retention each one uses, and whether a second run gives the same numbers.

In Kepler, the room was read in full when it arrived and resolved into your ontology next to the three earlier deals, their final models, IC memos and two years of post-close packages. Your definitions of gross margin and net revenue retention are part of the ontology, so code computes each metric for the target from the room’s customer cube and for the prior deals from their own models, every input cited to a cell. The change-of-control question returns every contract that has one, each clause quoted from its page. Listed competitors’ margins come from their filings, on each filer’s own reported figures. Where the CIM’s revenue differs from the audited financials in the room, both appear with the gap flagged. The screening memo drafts in house style, and an agent watches the room, updating the analysis and telling the team when new files or Q&A answers change a number.

After close, the same records carry into monitoring, so the KPIs in the first quarterly review use the definitions from the deal model. For more on the diligence side, see how to review a CIM with AI.

Setup, security and cost

V7 Go is scoped in an annual agreement priced on platform, users and pages processed, per V7, and V7 offers private deployment. Implementation by its solutions engineers is included.

Kepler has the public record live on day one. In the first week we connect your sources read-only, read every document and build your ontology, and our engineers stay on site as new use cases come up; anything a second firm would need goes back into the platform. We hold SOC 2 Type II, comply with GDPR and CCPA, isolate each firm’s data at the identity layer, run one access check on every read with walls between deals, log every query and calculation, and deploy in our cloud or yours (security).

When to choose Kepler

• One model should serve research and deals. Public-markets research, screens and earnings work run on the same model of the firm as diligence, monitoring and LP reporting.
• You need the complete set. Every contract with a change-of-control clause, every lease with a co-tenancy provision, counted and quoted.
• Your definitions should hold firm-wide. Net revenue retention, adjusted EBITDA and customer concentration defined once in the ontology, the way your IC defines them, and applied to every deal and report.
• Public comps are part of the work. Listed competitors and their filings, already read, next to the data room.
• The numbers go to an IC. Each figure reproduces, with its formula and source line, as reported or as restated.
• Monitoring follows the deal. The ontology from diligence carries into post-close packages, covenant tests and LP reporting.
• Agents should run overnight. On new uploads, Q&A answers, monthly packages and filings, acting as the person who set them up.
• Your team works in many tools. Excel, PowerPoint, Word, Claude, ChatGPT, Slack, or your own systems through our API.

When V7 Go may fit

• An insurance or real estate team automating submissions, claims or lease abstraction.
• A deal team that wants one specific workflow, such as CIM screening or DDQ completion, configured by a vendor’s solutions engineers and priced by page volume.

Next steps

For the wider field, see the best AI tools for private equity due diligence and the best AI portfolio monitoring tools. Kepler vs Hebbia covers the document-grid approach, and how private equity firms use AI in due diligence walks through the workflow.

Start at app.kepler.ai, or book a demo and we’ll run one of your data rooms against your prior deals.

Frequently asked questions

What is the difference between Kepler and V7 Go?

Kepler builds one living model of an investment firm from every document it holds plus the public record already loaded, reads every page at ingest, and computes every number in code with a citation to the exact page, line or cell. V7 Go is a platform of document workflows and agents for private markets, insurance and real estate, built on a Context Graph of a firm's files and priced by the pages it processes.

Is Kepler a good V7 Go alternative for due diligence?

Yes, especially when diligence should draw on the rest of the firm. Kepler reads the whole data room in full, links it to your prior deals, portfolio and house definitions, and returns complete lists, such as every contract with a change-of-control clause. Public comps come from SEC filings already loaded, and the IC memo drafts in your format with every figure computed and cited.

Does Kepler cite its sources like V7's visual citations?

Yes. Every figure opens the exact page, line or cell it came from, and a calculated figure opens a small sheet with its formula and each cited input. Kepler checks every citation against its source while the answer is written and removes one that doesn't match, and when two sources disagree it shows both.

Can Kepler monitor portfolio companies after close?

Yes. The ontology built during diligence carries into monitoring, so the KPIs you track after close use the definitions in your deal model. Kepler reads each monthly package, board deck and compliance certificate as it arrives, agents flag what moved with the exact quote, and the quarterly review drafts in your template.

How quickly can Kepler work on our documents?

Public company data is live on day one. In the first week our forward-deployed engineers connect your data rooms, SharePoint, inbox and CRM read-only, read every document in full and build your firm's ontology, then keep working with your team on site as new workflows come up.