Summary
- Perry AI has raised $2.2m in an oversubscribed seed round led by Haatch, with Antler, Arāya Ventures and Syncorp Ventures among the participating investors.
- Its platform maintains records of rights and obligations across private capital funds, using AI agents to assist legal reviews, drafting and transaction management.
- Customers reportedly manage more than $1.5bn in assets through the platform, although the company's claimed productivity gains have not been independently verified.
London legal technology startup Perry AI has raised $2.2m to expand a software platform that tracks contractual rights and obligations across private capital investments. The company is developing AI agents intended to help legal teams manage work that extends well beyond the completion of individual transactions.
The oversubscribed seed round, announced on 18 November, was led by Haatch, with participation from Antler, Arāya Ventures, Syncorp Ventures and international investors including Kelsey Chase, founder of investment analytics company Aumni. Perry intends to use the funding to develop additional capabilities and extend its commercial activity into the United States and Asia.
Founded by former funds lawyers Shashwat Patel and Vaneesa Agrawal, the company serves private equity, venture capital and other private capital organisations. Its software combines document analysis with a continuing record of legal commitments established through investment agreements, allowing later transactions and operational work to draw on information captured earlier.
Perry says funds using its platform manage more than $1.5bn in assets across over 1,000 portfolio companies. It also reports substantial improvements in the time customers spend on legal tasks, although those claims remain company-provided figures rather than independently controlled performance measurements.
Investment agreements create long-running obligations
Private capital transactions involve documents establishing rights and responsibilities that can remain relevant throughout the life of a fund. These may concern investor reporting, governance arrangements, future investment decisions or commitments made to particular counterparties.
The information is frequently distributed across agreements completed at different stages of a fund’s development. When a new transaction begins, legal teams may need to revisit earlier documentation to establish whether existing commitments affect the terms being considered.
Document management systems can store agreements and make them searchable, but locating a file is not the same as identifying every obligation it contains or understanding how that obligation interacts with later agreements.
Perry’s platform is intended to maintain a structured record of those relationships. Its AI agents examine legal documents, identify relevant provisions and assist with drafting and reviews using the fund’s existing contractual positions.
That approach could reduce repeated examination of historic material, particularly where similar legal questions arise during successive investments. However, its reliability depends on whether the original obligations have been interpreted correctly and whether the record remains accurate when agreements are amended.
An omitted provision could affect a future decision, while an incorrectly classified obligation might cause legal teams to overlook an approval requirement or rely on an inaccurate understanding of a counterparty’s rights.
The platform must therefore provide information that professionals can review and verify, rather than assume that automated extraction removes the need for legal judgement.
Specialised agents move beyond document summaries
Generative AI has made it easier for software to summarise legal material and produce initial drafts, but those functions are not necessarily sufficient for work involving several connected agreements.
Private capital teams may need to establish how provisions operate across different documents, whether particular consent rights apply or how an earlier commitment affects a new investment.
Perry is attempting to address these tasks through a continuing information structure rather than treating every document as a separate request. The company describes this as maintaining institutional memory at the level of the investment fund.
Its software supports activities involving fund formation, transaction execution, portfolio management and exits. The company says some expanded portfolio management capabilities are already available, with further development planned following the financing.
Among the workflows identified are tracking most-favoured-nation provisions, managing reporting commitments and coordinating information relating to co-investment allocations.
Those are specialist activities where the value of software depends on the accuracy and completeness of its underlying records. An AI system capable of producing convincing text does not necessarily possess the legal context required to interpret complex agreements reliably.
The founders’ professional backgrounds provide relevant expertise, but customers will still need evidence that the platform’s outputs are suitable for their individual circumstances and the legal obligations governing their work.
Customer performance figures require qualification
Perry reports that customers work 90% faster on selected legal tasks and that deal signing periods have fallen from 84 to 13 days in live transactions using the platform.
These results are significant if they can be reproduced, but the company has not provided an independently controlled comparison establishing how much of the improvement is attributable to the software.
Investment transactions vary considerably in complexity, negotiation requirements and the number of organisations involved. Some delays arise from commercial decisions or external approvals rather than the legal work that an automated platform can influence.
Assessing productivity therefore requires distinguishing the time spent reviewing and preparing documents from the overall period needed to complete a transaction.
Customers must also consider the work required to configure the service and verify its outputs. A platform that accelerates document review may still need substantial initial preparation to organise legacy information and establish appropriate access permissions.
Perry says customer information is not used to train its models or shared between clients. Those assurances address important confidentiality concerns, although prospective customers will need to examine the service’s security controls and contractual arrangements before introducing sensitive transaction material.
Expansion introduces legal and technical complexity
The company plans to develop the platform across the full investment lifecycle while expanding internationally. Private capital activity is global, but legal documentation and professional requirements differ between jurisdictions and transaction structures.
Supporting new markets will require the software to accommodate different contractual practices and forms of legal documentation without assuming that provisions operate identically in every location.
Competition is another consideration. Established legal software providers and general-purpose AI companies are developing products capable of analysing documents and assisting professional work.
Perry’s differentiation rests on its specialised understanding of private capital workflows and its ability to retain contractual context across multiple transactions. That may create value beyond individual document searches, although the commercial advantage will depend on sustained customer adoption and reliable results.
The deeper an organisation embeds such software into its processes, the more important data portability and service continuity become. Funds need to understand whether they can export their records and continue accessing historical commitments if their technology requirements change.
Investors are backing Perry at a point when specialist AI applications are attracting funding across professional services. The $2.2m round will support product development and expansion, but the next evidence of commercial progress will come from customer retention, the quality of the software’s legal records and independently demonstrable improvements in operational work.
For private capital organisations, automated assistance may reduce the work required to identify and manage existing commitments. The ability to preserve a reliable record of those obligations, rather than merely generate legal text quickly, will determine the usefulness of Perry’s approach.












