AI Consulting for Private Credit Funds
WorkWise Solutions is an AI consulting firm for private credit funds, direct lenders and BDCs. We pick the credit workflows worth handing to AI first, set the rules for borrower data and MNPI, and put Claude, ChatGPT or Microsoft 365 to work on your own credit documents. You can keep us on as your AI advisor, from $10,000 a month.
A credit fund runs on documents: credit agreements, quarterly borrower packages, compliance certificates and memos for the investment committee. Every number in them traces back to a page, so AI can do much of the reading while an analyst checks the result.
A lean credit team needs a practical AI strategy with guardrails for borrower data, and a plan that is manageable for the people it has. Specialist software can come later, once you know which work it has to hold.
We are not tied to any one AI vendor: we set up Claude, ChatGPT or the Microsoft 365 you already use. We sell no licenses and take no referral fees, so our advice on tools stays neutral.
All of our work is delivered live online, and every price is published on our pricing page.
By Dr. Leigh Coney, Founder of WorkWise Solutions. Updated November 2026.
Sources: Lumonic and Armentum Partners, 2026: The State of Private Credit Technology (about 150 private credit lenders, March 2026), for ~50%; PwC Global Private Credit Survey 2026 (more than 120 credit portfolio managers, fielded January to March 2026), for 54% and 16%, which describe intentions and priorities, so neither is a usage rate; Proskauer's 2026 private credit survey (January 2026), for 46%, up from 35% a year earlier. More figures in our private credit AI statistics roundup.
What We Do for Private Credit Funds
Six kinds of work, each with a published price. Start with a Sprint or the Credit-Team Intensive, buy a setup only if the plan calls for one, and keep us on through the retainer if you want someone to own AI.
Find where AI pays first
The AI Readiness Sprint takes one to two weeks and costs $12,500 for funds of up to 20 people. You get a check of your workflows and data security, a platform decision across Claude, ChatGPT, Microsoft 365 Copilot and Gemini, your credit workflows ranked, and a 90-day plan. Funds of 20 or more take the $30,000 Comprehensive Discovery Sprint, three to four weeks with governance and a hands-on pilot included.
Set up the tools safely
Our Claude setup for private credit funds takes three to six weeks: the right plan and seats, read-only connections to your data room and document stores, and private skills (saved instructions Claude follows for a task) built on your own memo template, spreading layout and covenant definitions. It costs $25,000 to $40,000 by firm size, with licenses bought from Anthropic. On ChatGPT Enterprise or Microsoft 365 Copilot, the Secure Platform Rollout runs at the same prices.
Write the fund's AI rules
The AI Governance Framework takes one to two weeks: an AI use policy, a supervision approach, vendor checks, answers for investors' due diligence questionnaires (LP DDQs) and a disclosure review. It sets which borrower data may go into which tool and how MNPI stays with the deal team. SEC-registered advisers can take the $17,500 SEC-Exam-Ready version, which adds an exam file for each AI tool and a mock-exam walkthrough with your chief compliance officer; both are operational documents, not legal advice.
Teach the credit team
The Credit-Team Intensive is three 2-hour sessions over two weeks for up to 12 people, on your own redacted credit work: spreading, covenants, memos and portfolio risk. It costs $12,500, and the fee credits in full toward any engagement of $25,000 or more booked within 90 days. Partners who want the overview first can start with a 90-minute Executive Briefing for up to 50 people, at $7,500.
Help choose credit software
Before you buy covenant, monitoring or document review software, we build a test set from your own credit agreements and borrower reports, sit in the vendor demos and score them with you. No vendor pays us anything. Vendor evaluations come on demand with the Plus retainer, at $15,000 a month.
Have someone own it
The AI Operating Partner retainer keeps us on as your fund's AI advisor: weekly office hours, one build at a time, a monthly usage review and standing advice for $10,000 a month. Plus, at $15,000, adds a second build slot and vendor evaluations; Embedded, at $20,000, is a fractional Head of AI with a seat in your weekly pipeline meeting. Three-month minimum, then rolling quarterly.
What Credit Funds Use AI For First
PwC's 2026 survey of credit portfolio managers gives the test: "If a task is document grounded and verifiable, AI is usually ready now." Spreading, covenant tests and memo drafts pass it, because an analyst can check every figure against its page.
PwC found that firms using AI are "reviewing more deals with the same headcount." It also sets a condition: "real value depends on the strength of underlying data, well-defined and integrated workflows, and the continued application of human credit judgement."
No survey measures how many hours a direct lender spends on each of these tasks, so the Sprint measures yours before anything is bought. Our guide to where a private credit fund should start with AI sets out the first 90 days.
Borrower package intake and spreading
Quarterly packages arrive as PDFs and spreadsheets in each borrower's own layout. Claude or ChatGPT can map them into your spreading template (the fund's standard layout for a borrower's financials) and list every line it could not map, so the analyst checks the exceptions. J.P. Morgan, writing about its own machine learning, describes "reducing the time taken for data extraction, which would typically take around 45 minutes for one company, to just 30 seconds."
Covenant tests
Claude or ChatGPT can list each loan's covenants from the credit agreement and recheck a compliance certificate against them, with an analyst confirming each definition, since in PwC's words "contracts are bespoke." The assistant keeps no test history, so when tests must run across the whole book with history and alerts, covenant software such as Lumonic, Allvue, Chronograph or 73 Strings becomes the record. Our guide to Claude for covenant monitoring shows the setup.
Credit memo first drafts
A house memo on private borrowers usually works in Claude or ChatGPT with your own template (see Claude for credit memos), and the analyst checks every figure before the investment committee sees it. PwC lists "investment committee memo generation" among the "more nascent but fast advancing areas." Anthropic's page on credit underwriting for commercial bankers puts the split this way: "Claude pulls spreads, runs ratios, and structures the writeup while you own the credit decision."
Portfolio monitoring, once the first three run
Monitoring comes after the first three, because it runs on clean spreads and a covenant list. When many people and your auditors rely on one record, that record belongs in monitoring software such as S&P Global's iLEVEL, Allvue or Chronograph, and Claude or ChatGPT reads its exports. Lumonic says its connector (a link that lets Claude read another system) gives "read-only access to periodic financials, covenant status, reporting status, and tracked KPIs."
Where Borrower Data Goes
The first thing a credit fund asks about AI is where borrower data goes. The answer depends on the plan you buy and the settings chosen on day one, so we put both in writing before anyone uploads a file.
Business plans do not train on your data by default. That covers Claude Team and Enterprise, ChatGPT Business and Enterprise, and Microsoft 365 Copilot. In Anthropic's words: "By default, we will not use your inputs or outputs from our commercial products (e.g. Claude for Work, Anthropic API, Claude Gov, etc.) to train our models."
The feedback buttons are the exception. On Claude, a thumbs up or down sends that conversation to Anthropic, where it can be kept up to five years and used for training. Team and Enterprise owners can switch ratings off, so decide that in the setup.
Personal accounts are the real risk. The personal plans (Claude Free, Pro and Max; ChatGPT Free, Plus and Pro) can train on chats unless the user switches it off. The fund has no admin control over those accounts, so borrower files stay out of them.
Chats are still kept, under the plan's retention rules. On Claude and ChatGPT, deleted chats are removed within 30 days. Claude Enterprise lets you set your own retention period (30 days at the least) and keeps audit logs, and ChatGPT Business and Enterprise admins control retention too.
Records, if you must keep them. Among Claude plans, only Enterprise has the Compliance API, a feed that lets an archive such as Smarsh, Theta Lake or Proofpoint capture conversations as they happen, before anyone can delete them. ChatGPT Enterprise has a Compliance Logs Platform that holds data for 30 days, so an archive has to download it continuously.
MNPI stays with the deal team. Material nonpublic information (MNPI) from a borrower belongs only with the people on that deal. We set up a project (a shared workspace) for each borrower or deal, open only to the group working on it, plus a written rule for public-side staff on what they may never open or paste.
Microsoft 365 Copilot shows each person what they can already open. In Microsoft's words, "Microsoft Copilot only surfaces organizational data to which individual users have at least view permissions." A credit file shared too widely in SharePoint becomes easy to find, so tidy file permissions before Copilot goes live.
More detail, plan by plan, is in our guide to whether Claude is safe for confidential deal data.
How a First Engagement Runs
A call with a partner or the COO
Thirty minutes on your team size, the systems your credit team uses, the AI plan you are on and where the hours go. If AI is not the answer, we say so on that call.
The Sprint or the Credit-Team Intensive
We interview the team, read a sample of your credit documents and look at the tools already in use, personal accounts included. You get a platform decision, ranked workflows and a 90-day plan. A team that already has its tool can start with the Credit-Team Intensive instead, on its own redacted spreads, covenant tests and memos.
Claude setup or the Secure Platform Rollout
We set up Claude for credit work, or ChatGPT Enterprise or Microsoft 365 Copilot through the Secure Platform Rollout, in three to six weeks: plan and seats, borrower files walled off by deal, read-only connections, and 30 days of support afterward. If you take the AI Governance Framework too, the fund's AI rules are signed in the same weeks.
The retainer, or a clean handover
The AI Operating Partner retainer keeps us on from $10,000 a month: weekly office hours, one build at a time and a monthly usage review, with a three-month minimum. If you would rather run it yourselves, we hand over written instructions and step back.
Prices
The retainer is listed first because most of the value comes after the setup, when the credit team uses AI every week. The fixed-price projects are the way in.
| Engagement | Price | What you get |
|---|---|---|
| AI Operating Partner, Core | $10,000 a month | Weekly office hours, one build slot, the monthly usage review, standing advisory and the LP DDQ answer bank |
| AI Operating Partner, Plus | $15,000 a month | Core, plus a second build slot, a monthly working session per team, a quarterly new-hire session, and connector and vendor evaluations on demand |
| AI Operating Partner, Embedded | $20,000 a month | A fractional Head of AI: a seat in the weekly pipeline meeting, live-deal support, roadmap ownership and a quarterly appearance at the investment committee |
| AI Readiness Sprint | $12,500 for firms of up to 20 people | One to two weeks: a workflow and data-security assessment, a platform decision, ranked use cases and a 90-day plan |
| Comprehensive Discovery Sprint | $30,000 for firms of 20 or more | Three to four weeks, with governance and a hands-on pilot included |
| Credit-Team Intensive | $12,500 | Three 2-hour sessions over two weeks for up to 12 people, on your own redacted credit work |
| Claude Setup for Private Credit Funds | $25,000 (1 to 10 people), $30,000 (11 to 25), $35,000 (26 to 60), $40,000 (61 to 150) | Three to six weeks: plan and seats, read-only connections, private skills on your own templates, then 30 days of support |
| Secure Platform Rollout (ChatGPT Enterprise or Microsoft 365 Copilot) | The same four prices by firm size | Three to six weeks, then 30 days of support |
| AI Governance Framework | $9,500 | One to two weeks: an AI use policy, a supervision approach, vendor checks, LP DDQ answers and a disclosure review |
| SEC-Exam-Ready AI Governance | $17,500 | Two to three weeks, for SEC-registered advisers: adds an exam file per AI tool and a mock-exam walkthrough with your chief compliance officer |
Retainer prices cover firms of up to 25 people: add $2,500 a month for 26 to 60 people and $5,000 for 61 to 150, with larger firms by proposal. Three-month minimum, then rolling quarterly; annual prepay earns one month free, and Embedded is capped at three firms at a time.
The Sprint and Credit-Team Intensive fees credit in full toward any engagement of $25,000 or more booked within 90 days. Licenses are bought from the vendor and not included. Prices from our pricing page, November 2026.
Who Else Could Do This, Compared Honestly
A boutique advisor is one of six ways a credit fund can get AI help. Each suits a different fund, and we are the wrong choice for some of them.
| Option | Suits a fund that | Published price | Limits |
|---|---|---|---|
| Large consultancies (the Big Four and similar firms) | Runs a large, multi-year program where AI sits beside audit, valuation or operations work | No published price | Staffed and priced for large programs, so a lean fund is a small client |
| Your credit software vendor's own services team | Has already chosen a monitoring, covenant or loan platform, such as Lumonic, Chronograph or Allvue, and needs it running | No published price; no vendor in these categories publishes one | Advice stops at the edge of their own product |
| Fund and loan administrators | Wants loan administration and investor reporting run outside the firm | No published price | They keep the loan record and run their own process; your analysts' spreading, covenant reading and memo work stays with you |
| A full-time head of AI | Runs AI across underwriting, monitoring and investor reporting at the same time | Posted base salaries: $140,000 to $195,000 plus incentive pay for a director AI role at Audax Private Debt (New York); $375,000 to $425,000 plus bonus for Ares's head of data and AI | Base pay only: each posting adds a discretionary bonus it does not publish; the largest managers split the role across several senior seats |
| A fractional head of AI | Needs a senior owner for a set number of hours a week | $8,400 to $14,000 a month at a typical 14 hours a week (Go Fractional marketplace data for a fractional chief AI officer) | Part-time, so someone inside the fund has to carry the work between sessions |
| A small AI consulting firm (for example WorkWise Solutions) | Wants a fixed-price first step, setup and training on its own credit documents, and an advisor on call | Retainer from $10,000 a month (Embedded, a fractional Head of AI, $20,000); Sprint $12,500; Credit-Team Intensive $12,500; Claude setup from $25,000 | You work directly with the founder, so there is key-person risk; no legal opinions; licenses bought separately |
Salaries are base pay as posted by each firm, and most of those listings have since closed. The fractional rate is Go Fractional's marketplace figure, read November 2026. WorkWise prices are from our pricing page.
Dr. Leigh Coney
WorkWise Solutions was founded by Dr. Leigh Coney, who holds a PhD in how humans interact with emerging technology. You work with him directly, and the work is delivered live online.
WorkWise has run 30+ engagements. WorkWise Solutions is an Anthropic Claude Partner, a Microsoft AI Cloud Partner and an OpenAI Select Partner. It sells no licenses and takes no referral fees, so the advice on which tool to use stays neutral.
AI Consulting for Private Credit Funds FAQ
Which AI consultants work with private credit funds?
Six kinds of help work with credit funds. Large consultancies suit big programs where AI sits beside audit, valuation or operations work. A software vendor's own team gets its platform running, and fund or loan administrators run the administration. A full-time or fractional head of AI owns the work from inside. A small AI consulting firm, for example WorkWise Solutions, sets up Claude, ChatGPT or Microsoft 365 on your own credit documents, from $12,500 for a first project or $10,000 a month on retainer. Our guide to AI consultants for private credit funds compares them.
How much does AI consulting cost for a private credit fund?
At WorkWise Solutions, the AI Operating Partner retainer costs $10,000, $15,000 or $20,000 a month for firms of up to 25 people, with a three-month minimum. Fixed projects: the AI Readiness Sprint is $12,500 for up to 20 people ($30,000 for the Comprehensive Discovery Sprint at 20 or more), the Credit-Team Intensive $12,500, a Claude setup $25,000 to $40,000 by firm size, and the AI Governance Framework $9,500. Licenses are extra. Credit software vendors publish no prices, so ask any provider for a fixed price in writing.
Where should a private credit fund start with AI?
Start with the document work your analysts already check: borrower package intake and spreading, covenant tests against the credit agreement, and first drafts of credit memos. PwC's Global Private Credit Survey 2026 gives the rule: "If a task is document grounded and verifiable, AI is usually ready now." Run those three on a business plan of Claude or ChatGPT, with an analyst checking each figure, and add portfolio monitoring once they work. A fixed-price first step: the $12,500 AI Readiness Sprint or the $12,500 Credit-Team Intensive.
Can AI read our credit agreements and borrower reports accurately?
Often yes, on your own documents, with an analyst checking the result. Claude or ChatGPT can pull definitions, covenant levels and borrower figures from a credit agreement or a quarterly package, and the analyst checks each term against its page. PwC's 2026 private credit survey names the hard part: "contracts are bespoke." So test any tool on your own agreements first, including how it reads your EBITDA definition and add-backs. When many agreements are reviewed at once, contract review software such as Litera Kira or Luminance fits better.
Is it safe to put borrower data and MNPI into Claude or ChatGPT?
On a business plan the fund controls, set up before anyone uses it, yes for most credit work. Claude Team and Enterprise and ChatGPT Business and Enterprise do not train models on your data by default; the personal plans can unless the user switches it off. Deleted chats are removed within 30 days on both, and Claude Enterprise adds custom retention and audit logs. For MNPI (material nonpublic information), give each borrower or deal its own project, open only to that deal team, with a written rule for public-side staff.
Should we buy private credit AI software or use Claude?
Start with Claude or ChatGPT on your own documents, and buy software when the work needs a system of record. Covenant software earns its place when tests must run across the whole book with history and alerts; Claude can recheck a compliance certificate but keeps no test history. Monitoring software fits when many people and your auditors rely on one record. A house credit memo usually works in Claude or ChatGPT with your own template. No vendor in these categories publishes prices, so test each one on your own files and get written quotes.
Does a private credit fund need a head of AI?
Most funds need one person inside who owns AI before they need a full-time head of AI. A full-time hire pays off once AI runs across underwriting, monitoring and investor reporting together. Posted base salaries show the cost: $140,000 to $195,000 plus incentive pay for a director AI role at Audax Private Debt, and $375,000 to $425,000 plus bonus for Ares's head of data and AI. Go Fractional's marketplace data puts a fractional chief AI officer at $8,400 to $14,000 a month. Embedded, the WorkWise retainer tier that works as a fractional Head of AI, is $20,000 a month.
Do you work with BDCs?
Yes, in the same way as other credit funds. A business development company (BDC) works from the same credit agreements, borrower packages, covenant tests and memos, so the same first steps apply: the AI Readiness Sprint or the Credit-Team Intensive, a setup on Claude, ChatGPT or Microsoft 365, then the AI Operating Partner retainer. Where the BDC or its manager also has public-side staff, the MNPI rule comes first. Questions about BDC rules go to your counsel; our work is operational, not legal advice.
Questions Small Credit Teams Ask
Each guide answers one question a lean credit team asks before its first AI project, with prices for a firm of 10 to 40 people.
What Is the Most Affordable Way for a Small Private Credit Fund to Use AI?
Business seats at $20 a month, a fixed-price setup or an advisor on retainer: what a lender of 10 to 40 people pays for each route.
Is AI Worth It for a Small Private Credit Fund?
The published results from lenders, banks and vendors, what they do not show for a fund of 10 to 40 people, and how to measure it yourself.
Can a Private Credit Fund Use AI Without an IT Team?
Yes: one owner inside, your outside IT provider for sign-in, and someone senior on call for the rules and the first workflows.
How Does a Small Credit Team Monitor 50 Borrowers Without Adding Headcount?
Collect, spread, test and write up: where a small team's hours go each quarter, and the first pass AI can take on each step.
How Do You Stop Chasing Borrowers for Financial Statements?
A reporting calendar from each credit agreement, reminders, a same-day package check and a late list, before you buy software.
Can ChatGPT or Claude Spread Borrower Financials?
Yes for clean statements, with every figure checked against its page. Where the assistants slip, and the routine that holds up.
Is It Safe to Put Borrower Financials Into ChatGPT or Claude?
The plan, the confidentiality clause in each credit agreement and your own rules decide it. What a small lender must check first.
How Can a Private Credit Fund Cut Quarterly Portfolio Review Prep From Weeks to Days?
Most of the weeks go to gathering, keying and formatting. The four steps AI can draft, and what your people still decide.
When Should a Private Credit Fund Move Off Spreadsheets?
Four signs a fund has outgrown the workbook, what to fix before buying anything, and the technology a small credit firm needs.
Should a Private Credit Fund Build Its Own AI Tools or Buy Them?
Configure first, buy second, build last: three routes compared on cost, upkeep and fit for a fund of 10 to 40 people.
Talk to Us About Your Fund
A 30-minute call with a partner or the COO: your team size, the systems your credit team uses, and whether a Sprint or the Credit-Team Intensive is the right first step. If neither is, we'll tell you.
Book a CallOr email Contact@workwisesolutions.org
Related Guides for Private Credit Funds
Where Should a Private Credit Fund Start With AI?
Spreading, covenant tests and the credit memo first, then monitoring: the first 90 days for a lean credit team.
How Should a Private Credit Fund Choose AI Software?
A scorecard, the questions to put to every vendor and a four-week selection on your own credit agreements.
Does a Private Credit Fund Need a Head of AI?
Hire, fractional or retainer: what each costs, what you get, and when a full-time hire pays off.
AI Consultants for Private Credit Funds
Large advisory firms, fund administrators, software teams, specialists and fractional help, compared on cost and fit.
Private Credit AI Statistics 2026
How many lenders use AI, where they apply it and how ready their data is, every figure with its source.
How to Set Up Claude at a Private Credit Fund
Set the data rule first, start with covenant tracking or monitoring, then prove it on real borrowers.
What Claude Costs a Private Credit Fund
Seats, Enterprise usage, setup and training, with year-one budgets for firms of 10 and 25 people.
Claude for Covenant Monitoring
Pull covenant terms from the credit agreement, compute headroom and flag breaches early across the loan book.
Who Should Own AI at Your Firm?
The one-owner rule, what the owner does, and whether you need a head of AI yet.