Beyond the Standard Report: Building Custom Views with InvestNext MCP

Contributions by

Share this article

In this article: We’re sharing five ways GPs can use AI and InvestNext MCP to explore their data beyond the boundaries of a standard report, from building custom project views and comparing periods to checking distribution readiness and uncovering patterns across their investor base.

There’s a familiar moment that happens when you’re looking for an answer in your data:

You know exactly what you want to know. You just don’t have a report built for it.

Maybe you need three projects instead of your entire portfolio. Maybe you want to compare this month to last month. Maybe the question cuts across funds, investor groups, or time periods in a way that doesn’t fit neatly into a standard export.

So you pull a report that gets you close. Then another. Then the spreadsheet opens.

In our conversations with InvestNext MCP customers, reporting has emerged as one of the clearest examples of what changes when AI can work directly with the data already inside InvestNext.

Instead of starting with “Which report has this?”, you can start with “What do I want to know?”

That opens up a much broader definition of reporting. A report can be something you ask for in the moment, a recurring analysis built around the way your firm operates, or a specific question about your investors that you’ve never had a reason to build a dashboard around.

Here are five ideas for exploring what that could look like.

Quick Glossary: A Refresher on AI Terms

If you’ve been following this series, these will look familiar.

  • MCP (Model Context Protocol): A standard that allows AI applications to securely connect to external tools and data sources. InvestNext MCP allows supported AI tools to work with information from your InvestNext account.
  • Claude: Anthropic’s AI assistant and the tool used in these examples.
  • Prompt: An instruction or question you give an AI assistant to complete a specific task.
  • Skill: A reusable set of instructions that helps Claude perform the same type of task consistently without rebuilding the workflow each time.

1. Build the Report You Actually Need: Custom, Ad Hoc Views

One of our early access customers put the reporting challenge simply: sometimes the standard options are “all or one.”

But the question you’re trying to answer might sit somewhere in between.

You might want to see five projects that share a strategy. Or compare a subset of active funds. Or pull information for a specific group without building and combining several separate reports.

With InvestNext MCP, the scope of the question can become the scope of the report.

What Claude can pull from InvestNext:

Project, fund, investor, commitment, position, transaction, and distribution information available to the user.

How to explore it:

Start with a prompt. Tell Claude exactly which projects you want included, what information you care about, and how you want the results organized.

This is particularly useful for questions that matter today but may never warrant their own permanent report.

If you find yourself asking for the same view repeatedly, turn those instructions into a skill so your team can recreate it consistently.

Try this in Claude:

Review [Project A], [Project B], and [Project C] in InvestNext.

Create a report that includes:

– Total commitments by project
– Number of investors by project
– Total distributions to date
– Most recent distribution date
– New commitments in the last 90 days

Include a total across all three projects and clearly label the as-of date for the data.

Do not include projects outside this list.

2. See What Changed: Period-Over-Period Analysis

Sometimes the most important number isn’t the number itself. It’s what changed.

A GP reviewing performance month after month may have two perfectly accurate reports sitting side by side and still need to do the analytical work of figuring out what moved between them.

That’s another place where customers saw an opportunity for AI.

Rather than reviewing every line manually, ask Claude to compare two periods and direct your attention to the differences.

What Claude can pull from InvestNext:

Current and historical investment, commitment, transaction, and distribution information available for the projects being analyzed.

How to explore it:

Define the periods you want to compare and, importantly, what counts as meaningful.

You may care about every change. Or you may only want to see movements above a certain dollar amount or percentage.

Claude can organize the comparison and surface the differences for review. Your team can then investigate the movements that actually deserve attention.

For a recurring monthly or quarterly review, this is a strong candidate for a skill.

Try this in Claude:

Compare activity across [Project Name(s)] for [Period 1] and [Period 2].

Show me:

– Changes in total commitments
– New investor activity
– Distribution activity
– Any other material changes visible in the InvestNext data

For each change, show the previous value, current value, and difference.

Flag changes greater than [$X or X%] at the top.

Do not speculate about why a number changed. Surface the change and the supporting data so our team can investigate it.

3. Check Before the Money Moves: Distribution Readiness

A distribution report can tell you what should happen.

Sometimes the more useful question is: What could prevent it from happening?

Before a distribution goes out, there may be details worth checking across the investors receiving it. Missing payment information is one example. Depending on your process, incomplete documents or other gaps in the investor record may also warrant attention.

Instead of finding those exceptions one investor at a time, AI can help turn readiness into a report of its own.

What Claude can pull from InvestNext:

Investor, position, payment, and relevant document or account information available in InvestNext.

How to explore it:

Give Claude the project or distribution you’re preparing and define the conditions your team wants checked.

The goal isn’t to have AI decide whether a distribution should proceed. It’s to create an exception list your team can review before taking action.

This becomes especially useful when you define the same readiness criteria before every distribution. Those instructions can become a reusable skill and a consistent pre-distribution checkpoint.

Try this in Claude:

Review the investors associated with [Project Name] in preparation for our upcoming distribution.
Identify any investor records that may need attention before the distribution is processed based on the information available in InvestNext.

Check for:

– Missing or incomplete payment information
– Missing required tax documentation
– Incomplete subscription documentation
– Any other incomplete information relevant to the distribution that is visible in InvestNext

Return only the investors who need review.

For each, include the investor name, the item that needs attention, and the relevant information currently on file.

Do not determine whether the distribution should proceed. Give our team the exception list to review.

4. Slice Your Investor Base Differently: Segmented Reporting

Project-level reporting answers plenty of questions.

Investor relationships don’t always fit neatly inside project boundaries.

A firm might want to understand investors from a particular vintage, compare participation across different offering types, or look specifically at longtime investors across several funds.

Those questions require a different lens on the same underlying information.

What Claude can pull from InvestNext:

Investor profiles, investment and commitment history, project participation, positions, and other relevant attributes available in InvestNext.

How to explore it:

Start by defining the group you want to understand.

Instead of asking for a project report, describe the investor population you care about and then ask the question.

For example, you could look at investors who first joined the firm in a particular year and see how their participation evolved. Or isolate investors in a specific offering and understand their commitment patterns across subsequent opportunities.

The important shift is that the segment comes first.

Try this in Claude:

Identify investors in InvestNext who [DEFINE SEGMENT, such as “made their first investment with us in 2024”].

For this group, return:
– Investor name
– Initial investment date
– Initial commitment amount
– Total number of investments
– Total historical commitments
– Most recent investment date
– Current active investments

Then summarize the group as a whole, including average initial commitment, average total commitments, and average number of investments per investor.

Only use information available in InvestNext. Clearly note any fields that are unavailable rather than estimating them.

5. Ask Bigger Questions About Your Investor Base: Benchmarking Analytics

Once you stop thinking of reporting as a fixed set of fields, another possibility opens up.

You can start asking questions about patterns.

How does average check size change as an investor’s relationship with your firm gets longer? How many investors participate in multiple funds? How concentrated is your capital among your largest LPs? Do commitment patterns look different between newer and longtime investors?

These aren’t necessarily reports you need every week. They’re questions that can help your team better understand the capital and relationships behind the portfolio.

What Claude can pull from InvestNext:

Investor tenure, investment and commitment history, participation across projects or funds, and other relevant investor and position information.

How to explore it:

Start with one specific hypothesis or question.

The narrower the question, the easier it is to define the population, calculation, and output you actually need.

And ask Claude to show its work. For analytical questions, the summary matters, but so does being able to see the underlying groups and calculations your team should verify.

Try this in Claude:

Analyze our investor history in InvestNext to understand how investment behavior differs by investor tenure.

Group investors based on how long they have invested with our firm:

– Less than 1 year
– 1 to 3 years
– 3 to 5 years
– More than 5 years

For each group, calculate:

– Number of investors
– Average commitment amount
– Median commitment amount
– Average number of investments per investor
– Total historical commitments

Then summarize any notable differences between the groups.

Show the underlying calculations and sample sizes. Do not infer why the differences exist or make assumptions about investor behavior beyond what the data supports.

From Reports to Questions

Traditional reporting starts with structure.

Here are the fields. Here are the filters. Here is the report you can run.

But your questions don’t always arrive that neatly.

Sometimes you need three projects, not all of them. Sometimes you want to understand what changed rather than what exists. Sometimes the useful view cuts across funds, vintages, investors, and time periods in a combination nobody had a reason to build in advance.

That’s what makes reporting an interesting place to experiment with InvestNext MCP.

The goal isn’t to create more reports. It’s to reduce the distance between having a question and having the information you need to think about it.

Start with the question your team would normally answer by exporting data, combining spreadsheets, or asking someone to pull numbers for you.

Then see what happens when you can just ask.

Start somewhere. Stay curious. Move with purpose.

A Note on Accuracy

Flexible reporting still requires oversight. Review AI-generated analysis before using it for investor communications, financial decisions, distributions, board materials, or other consequential workflows. Check important figures against the underlying InvestNext data, particularly when calculations, comparisons, or multiple data sets are involved.

AI can make it much faster to get from your data to an answer. Your team still decides whether that answer is accurate, relevant, and ready to use.

Already an InvestNext client? Contact your account manager about getting access to InvestNext MCP.

Not a current client but interested in learning more about InvestNext and InvestNext MCP? Schedule a demo today.

Author

  • Kaylee has a diverse background in software and technology, primarily focused on community building and software empowering small to mid-sized businesses to succeed and scale. At InvestNext, she leads client profiles research, product storytelling, and leverages marketing insights to drive content initiatives for the company.

    View all posts

Contributors

Schedule a Demo

Connect with our product experts to get a personalized view of InvestNext

Share this article