Zuva Bags $15.75m, Sets Out ‘DocAI’ Strategy

Zuva, the spin-out from Kira Systems, has completed a $15.75m (CAD $20m) Series A funding round, with Insight Partners leading, and with Litera’s owner Hg also making a minority investment. The corporate-facing doc analysis company, headed by Noah Waisberg, also set out its strategy, (see following story for in-depth AL TV interview), that will initially focus on its new product, DocAI.

The company explained that DocAI is an ‘API-driven platform that enables developers to implement AI features into their applications, without requiring AI development from the ground-up’.

It can classify documents and extract key clauses written in non-standard natural language from business contracts and documents, and comes with over 1,200 pre-trained AI models. Users can also create their own models without requiring data science or machine learning development knowledge, and leverage Zuva’s cloud platform for their hosting needs, or opt to host it on their own infrastructure on Kubernetes.

Waisberg told this site that the patents for the NLP tech used by Zuva are shared with Litera, which bought Kira after ten years of growth for an undisclosed sum. Waisberg also told Artificial Lawyer that there was a real possibility that Zuva could in the future provide its tech capabilities back to Litera and its acquisition, Kira, as they continued to develop their NLP capabilities.

And as noted in August, Waisberg will also maintain a close relationship with Litera, providing strategic input when needed. Insight also has a shareholding in Litera after the Kira deal.

So, given the above, it would seem that the two businesses will remain fairly closely linked, even if Zuva is an entirely separate business from a corporate and revenue generating standpoint. It will also be based in Toronto, Canada, where Kira has been based since it started.

In terms of where the Insight money came from – which announced a $50m investment in Kira back in 2018 – Artificial Lawyer asked if this was part of the same money from then, but that has now been re-allocated to Zuva? Waisberg replied: ‘Money is a fungible thing. Insight had the opportunity to take money back or roll it forward.’

In this case it looks like they have rolled some of it forward.

As it stands today, Zuva is staffed by about 35 people from Kira, with a new hire joining this week to help with HR. A notable number of those who came from Kira are involved in technical matters, i.e. related to product development, machine learning and NLP.

In the video, (see next story), Waisberg also explained that the goal now is to cover potentially all of a corporate’s doc analysis needs, ranging from legal – as one would expect – to invoices, to finance and banking documentation, to leases, company minutes, and even corporate press releases.

Zuva will offer the ability to first find and categorise these documents across a company, then using its ‘smart fields’ offer analysis of the contents – although that is a work in progress for some of the document types. For now they have many hundreds of smart fields that one would expect to see in commercial documents that can be applied across a wide range of contracts and other company paper. So, that should get them a long way.

In terms of the business approach, there are several paths they may take. One way is to partner with consultants or other organisations that are handling complex unstructured data and document intelligence projects for corporates, where Zuva can be the tech solution to the problem of the client.

Another path, as noted above, is to provide Zuva’s NLP capability to other tech companies and become a kind of ‘software component’. Waisberg suggested that you could see Zuva as a kind of Plaid or Twilio, in that they can become part of a much broader offering.

A third path is to work directly with larger corporates that have enough spare internal capacity to manage the kind of large scale doc analysis projects that Zuva will be suited to.

Peter Sobiloff, Managing Director at Insight Partners, who will also join the board at Zuva, said: ‘We are eager to continue to partner with Noah, Alexander and the team coming over to Zuva.

‘This is a team with tremendous machine learning expertise. We continue to see an opportunity for the technology they have built and look forward to supporting their future growth in the corporate market.’

Zuva has already bagged some clients, including MNTSQ and Vigilant AI.

But, for a more in-depth insight into Zuva, please see the AL TV video interview in the following Artificial Lawyer article.

P.S. just so you can get a sense of what Zuva is doing, here is a list of contract and document types they are aiming to help with – first to identify, then in many cases to get deeper into the documents to extract key data.

Contract Types

‘The classifier will identify the following types of contracts:

  • Business Transaction Agt – Agreements entered into in connection with (or ancillary to) a business transaction. Includes purchase agreements, subscription agreements, underwriting agreements, escrow agreements and letters of intent.
  • Credit-Related Agt – Agreements entered into in connection with (or ancillary to) a loan transaction. Includes credit agreements, commitment letters, guarantees and promissory notes.
  • Distribution Agt – Agreements pursuant to which a party agrees to distribute goods or services for another party. Includes distribution agreements, franchise agreements, wholesale agreements and reseller agreements.
  • Employment-Related Agt – Agreements entered into in connection with (or ancillary to) an employment relationship. Includes employment agreements, change in control agreements, incentive compensation agreements and collective bargaining agreements.
  • Equipment Lease – Agreements pursuant to which one party leases equipment to another.
  • Governance Agt – Agreements with respect to the governance of an entity, including shareholders’ agreements, partnership agreements, operating agreements and joint venture agreements.
  • Indenture – Agreements between issuers of debt securities and trustees. Includes high yield indentures, investment grade indentures and agreements ancillary thereto.
  • Insurance Agt – Agreements with respect to insurance, including insurance policies and reinsurance agreements.
  • Intellectual Property Agt – Agreements governing the use or transfer of intellectual property. Includes license agreements, royalty agreements, intellectual property transfer agreements and development agreements.
  • Litigation-Related Agt – Agreements pursuant to which parties agree not to take litigious action against one another, to settle disputes between them or to indemnify each other in specific circumstances.
  • NDA or Non-Compete Agt – Nondisclosure or confidentiality agreements whereby a party agrees not to disclose information confidential to the other party. Also includes non-competition, non-confidentiality and non-disparagement agreements.
  • Real Estate Agt – Agreements entered into in connection with (or ancillary to) a real estate transaction. Includes real estate purchase agreements, real estate leases and related assignments and guarantees.
  • Service Agt – Agreements whereby a party agrees to provide services to another party. Includes service agreements, consulting agreements, engagement letters, financial services agreements, management services agreements and transition services agreements.
  • Structured Finance Agt – Agreements that govern structured finance or derivatives transactions. Includes ISDAs and repurchase agreements.
  • Supply Agt – Agreement pursuant to which a party agrees to supply goods to another party. Includes supply agreements, quality agreements and warranties.

Non-Contract Types

The classifier will identify the following non-contract types:

  • Accountant Consent – Consents by accountants to the use of their firm name in public disclosure documents
  • Banking Document – Forms or documents used in banking. Includes deposit slips, loan forms, pension forms and powers of attorney.
  • Retail Receipt – Documents evidencing payment.
  • Court-Related Document – Statements or documents issued by or submitted to a court. Includes arbitration rulings, complaints, court opinions and pleadings
  • Disclosure Document – Disclosure documents of a company. Includes information statements, offering memoranda, prospectuses and statements of executive compensation.
  • Due Diligence Material – Documents setting out the parameters of due diligence. Includes checklists and request lists.
  • Financial Statements – Financial statements of a company.
  • Invoice – Invoices or bills.
  • Lease Application or Abstract – Lease applications and abstracts.
  • Manual – Documents setting out instructions or procedures, such as an instruction manual.
  • Minute Book Material – Minute book material, including director or shareholder resolutions and minutes of a meeting.
  • Organizational Chart – Diagrams showing the structure of an organization.
  • Policy – Documents setting out principles and guidelines of an organization. Includes board compensation policies, corporate governance policies, health and safety policies and privacy policies.
  • Press Release – Official statements issued to press about a particular matter.
  • Request for Proposal – Documents that solicit proposals through a bidding process.
  • Rules – Sets of explicit regulations or principles governing conduct, such as competition or board game rules.
  • Tax Document – Forms submitted to governmental bodies regarding tax matters.
  • Wire Transfer – Forms containing wire transfer information.’