The first thing to fail was Excel. Not because Excel is bad, but because private credit became very, very good at producing workbooks. Every loan arrived with its own agreement, covenant definitions, borrower reports, amendments, cash flows and reporting calendar. Then funds added leverage facilities with another set of tests. What looked like a portfolio from the boardroom looked like an archipelago from the operations desk.
Sumit Gupta and Vishal Soni knew the feeling. The former Deutsche Bank credit professionals founded Oxane Partners after confronting the same awkward truth in their investment careers: a pure software vendor might understand databases but not the deal, while an adviser might understand the deal but leave behind another spreadsheet. Their answer was neither category on its own. It was a platform staffed by people who could argue with the data.
That hybrid has become a sizeable, privately held business. Oxane says it now supports more than 100 institutional clients and over $1.5 trillion of notional, including 23 of the world's 30 largest investment banks and 13 of the top 30 private-debt funds. Its workforce has expanded beyond 950 people across London, New York, Gurgaon and Hyderabad. In July 2026, growth investor TA Associates agreed to back the next phase. The price was not made public. The intent was: more technology, more AI, more talent and a wider global reach.
The product is a pane of glass. The work is underneath it.
Oxane Panorama is the visible center of the company: a modular system for portfolio monitoring, leverage management, valuations, enterprise data, loan servicing, deal pipelines and AI-assisted work. It covers corporate credit, fund finance, asset-based lending, risk-transfer trades, securitized loans, whole-loan portfolios, real estate and project finance. That list sounds like a software menu until you consider what the modules are replacing.
A borrower sends a financial package. Someone must extract the numbers, map them to the lender's definitions, validate the result, calculate covenants, flag exceptions and preserve the trail for the next reporting date. A fund using borrowed money repeats the exercise for its own lenders, whose templates and eligibility rules may differ. A valuation committee then wants to know why a mark moved. An investor wants a report. A regulator wants evidence. Each request touches the same underlying loan but often lives in a different system.
The Panorama loop
Panorama tries to keep that loop in one governed environment. Its AI layer extracts agreement terms, spreads financials, ingests unstructured documents and helps draft commentary. The more important design choice is less fashionable: human review remains in the workflow. Private-credit agreements are negotiated objects. Two covenants bearing the same name can behave differently after baskets, holidays and amendments are applied. Automation can move the first pass; accountability still needs an owner.
“Don't work on data. Put data to work.”Oxane's pitch for AI in private credit
The spreadsheet did not lose. It got promoted.
Oxane's public client stories are anonymous - normal in institutional finance - but unusually specific about the before state. One global investment firm, active in more than 20 countries and managing over $150 billion, ran billions of leverage commitments through bespoke lender templates and manual borrowing-base calculations. The result was slow compliance reporting, a risk of delayed funding and no clean cross-portfolio view.
After moving the operation onto an integrated setup, the firm cut lender-reporting turnaround by 50 percent and data-management overhead by 80 percent. Another alternative credit manager, previously consolidating monthly Excel files from multiple originators and servicers, reduced risk-analysis turnaround from days to minutes with one-click reporting. For three global investment banks, the trigger was similar: growing illiquid-credit books had made manual portfolio systems too fragile for the volume and volatility.
The lesson is not to ban the spreadsheet. Investment teams will always want a flexible scratchpad. It is to stop asking a scratchpad to be the source of truth, workflow engine, audit log and reporting factory at the same time. In Oxane's model, Excel becomes an interface or an export rather than the institution's memory.
Software with a credit desk attached
Most enterprise vendors sell a system and try to minimize the humans required to run it. Most outsourced service providers sell humans and treat technology as supporting furniture. Oxane sits between them. Its “Platform x People” model pairs proprietary software with teams of financial analysts, technologists and data specialists. The company can configure a workflow, operate parts of it, perform independent valuations or act as facility, security, verification, master or backup servicer.
That positioning makes sense for customers whose portfolios are both large and irregular: global banks, direct lenders, asset managers, insurers and limited partners. They are buying control over bespoke assets rather than a cheerful off-the-shelf dashboard. Oxane makes money through enterprise software and technology-enabled services, configured around modules, asset classes, integrations and the amount of operational support involved. There is no public price card because this is closer to fitting a trading floor than downloading an app.
The competitive set changes with the buyer's headache. Allvue, eFront, SS&C Advent and Dynamo compete for investment-management workflows. Versana and Cardo AI work in adjacent credit-data and monitoring territory. Document specialists attack ingestion. Administrators offer people. And the incumbent is still a home-built stack of Excel, email, data warehouses and heroic employees who know which tab not to touch.
Where Oxane earns its keep
Many asset classes, bespoke terms, repeated reporting, lender leverage, hard-to-value positions and a genuine need for auditability.
Where the math gets weaker
A small, standardized loan book; clean data already living in one system; light reporting; or a team determined to own every model and workflow in-house.
What changed the buyers' minds
The pattern across Oxane's case studies is not a sudden affection for digital transformation. It is operational pain crossing a threshold. A manager launches a new strategy and wants infrastructure ready before its first close. A treasury team adds facilities until reporting threatens funding speed. An investment bank's illiquid book grows across jurisdictions and asset types. A structured-credit fund needs independent marks that satisfy investors, auditors and AIFMD rules. The old method works - until volume, scrutiny or complexity arrives together.
Private credit's own success keeps moving that threshold forward. Banks and non-bank lenders are more intertwined; portfolio leverage can be material relative to a fund's assets. Investors ask for more frequent information. Valuations increasingly affect liquidity, financing and fundraising conversations, not merely the quarter-end ritual. Oxane's broad suite is a bet that these are not separate software problems. The asset, the leverage against it, its valuation and its reporting should share a data spine.
The bet also carries implementation risk. A unified platform is only as trustworthy as its definitions, ownership and exception handling. Historical data has to be cleaned. Bespoke clauses must be translated without sanding off their meaning. Upstream servicers and downstream accounting systems still need connections. If a team treats deployment as a technology purchase rather than an operating-model change, the unofficial spreadsheet will quietly return.
The part worth copying
Oxane's most portable idea is sequencing. Start with one painful, recurring workflow - monthly borrower reporting, a borrowing base or a valuation process - and establish a controlled data model underneath it. Assign owners to fields and exceptions. Keep the original document beside every extracted term. Automate validation before presentation. Make the audit trail part of the daily workflow, not a reconstruction performed when someone asks. Only then add dashboards and generative AI.
A second lesson is organizational. Put the credit expert and the technologist in the same design conversation. The former knows why the clause matters; the latter knows how to make the answer repeatable. Oxane was built on the founders' view that neither discipline could solve the problem alone. Twelve years of client growth suggest that the awkward middle is not a compromise. It is the product.
This approach will not fit every lender. Standardized portfolios may be served by narrower tools. Small funds may not generate enough reporting friction to justify an enterprise implementation. Firms with strong internal engineering and a genuinely consistent data model may prefer to build. And no platform can compensate for late borrower data, ambiguous legal terms or a committee unwilling to agree on definitions. The machine can expose an argument; it cannot settle one.
Still, the direction of travel favors Oxane. In 2025 the company collected awards across portfolio management, valuation and AI credit-risk analysis. In 2026 TA supplied growth capital, while Oxane pushed beyond direct lending into the wider field it calls “Private Credit+” - real estate, infrastructure, asset-backed finance, consumer credit, hard assets and securitized products. That label is marketing, but it describes a real convergence: more forms of private lending are becoming institutional, financed and data hungry.
The headline number is $1.5 trillion. The more revealing unit is one hour, or one fewer reconciliation, or one covenant caught before funding day. Oxane's business is built on making private credit less mysterious to the people already inside it. The loans can remain bespoke. The plumbing no longer has to be.