top of page
  • Writer's pictureStartU HQ


Klarity Tackles Laborious Contract Review with AI

Klarity is a cloud-based artificial intelligence powered contract review software. Legal Tech may not have the same buzziness of other tech sectors such as autonomous cars or VR, but contractual legal work represents a $280 billion market that is a huge time sink and pain point for businesses of all sizes and across all industries. I spoke with CTO and co-founder Nischal Nadhamuni as the company contemplates the next steps after its whirlwind journey through MIT’s delta v accelerator.

Automated Contract Review Born Out of a Harvard Law and MIT Connection

Andrew Antos (CEO & Co-Founder) points to his experience at a law firm, where he spent countless soul-sucking hours poring over contracts, as crystalizing his belief in the need for a more automated solution to contract review.

As Nischal Nadhamuni (CTO & Co-Founder) put it to me: “no lawyer wakes up and says ‘I want to review an NDA today’.” Having spent a summer working as a paralegal at a law firm in New York, this sentiment certainly struck a chord with me.

Klarity was born out of Bill Aulet’s legendary New Enterprises class at MIT Sloan. Andrew, a Harvard Law student cross-registered in New Enterprises, and Nischal, then a MIT computer science undergraduate, found themselves sitting next to each other and ended up talking for three hours about how the machine learning concepts Nischal was working with could help address the contract review problem.

Co-Founders Andrew Antos (CEO, left) and Nischal Nadhamuni (CTO, right)

Early Pivot to Exclusive Focus on Non-Disclosure Agreements

Contract review is a vast market with nearly limitless ways to segment the customer. When they first started pitching customers on their machine learning powered contract review product, Klarity focused on law firms with a product covering a broader range of contracts

“We knew that it took some warming up to having AI process documents, so we didn’t necessarily want to start with the biggest ticket item.” Nischal Nadhamuni, Co-Founder of Klarity

During their participation in MIT’s delta v accelerator last summer, they pivoted to focus on selling to in-house legal departments rather than law firms, with an initial product that automates the review of non-disclosure agreements (NDAs) exclusively. This focus was beneficial because NDAs are a problem due to their volume, not the individual value of each NDA. This combination makes businesses willing to test software on a task that was previously done solely by humans.

“NDAs are significantly complex to show off the full gamut of our analytical capabilities but at the same time they are not such risk heavy contracts that people would be unwilling to try our software on it.” Nischal Nadhamuni, Co-Founder of Klarity

Beyond the focus on NDAs, Klarity further narrowed their focus to enterprise software companies. The impetus for this concentration was the team’s observation of the most frequent contracts they saw other startups in the delta v accelerator requiring. Beyond the large number of NDAs enterprise software companies require over the course of their sales process, a second benefit to working with enterprise software companies is that most NDAs are signed on their customer paper because they try to speed up the sales process. This fact means that they must devote more resources to reviewing NDAs.

2017 delta v teams at the delta v Demo Day

Data Acquisition & Annotation a Huge Hurdle

A major challenge on the technical side was acquiring data to train the natural language processing neural nets. First, they had to find a large enough set of NDAs for a starting point for their data training set. Luckily they were able to scrape close to a million NDAs from public sources, such as SEC filings. This was only the first step, however; Andrew had to comb over thousands of the NDAs to identify specific clauses to build a robust training set for the machine learning powered review engine.

“[On annotation] we go to extreme granularity because we are not just saying what do we need to train our models on for today; its about the next year and the next two years, what are the pieces of information we need because data annotation is actually a pretty expensive process.” Nischal Nadhamuni, Co-Founder of Klarity

Their product can now identify and parse over 70 NDA clause types in and they are continually adding more clauses. Luckily, however, Antos now has help in annotating the contracts for the training set from lawyers.

A Klarity reviewed contract with annotations

Data Security Foundational to Product

“Security was the biggest focus for us from day one because we are dealing with highly confidential information.” Nischal Nadhamuni, Co-Founder of Klarity

Another technical hurdle for Klarity was ensuring the security of data on their platform. In many startups, security can be put on the backburner to a certain extent, but Klarity’s focus on sensitive contracts left air-tight data security as a non-negotiable. From the get go, Klarity incorporated end-to-end encryption and followed various standards for data security including ISO 27001 as well as recruiting a panel of security advisors.

Expanding Beyond NDAs and Enterprise Software without Diluting Depth of Analysis

There are other players in the legal document review space but their products either do not deliver the same level of insight as Klarity or deliver insights at a far slower pace due to increased reliance on human lawyers.

Therefore, Klarity is proceeding methodically as they look to expanding in terms of both type of contracts and industries outside enterprise software. The key determinant on the industry side remains large amounts of sales-side contracts that need to be reviewed spread across few contract types.

In terms of expanding beyond NDA’s, Klarity is moving up the sales process and has already begun development of a module that automatically reviews Master Service Agreements (MSAs) and Service Level Agreements (SLAs).

Talent Acquisition and Data Acquisition Focus for Internal Growth

Internally, Klarity’s primary focus for growth is talent acquisition followed by data acquisition. The engineering team is based in Bangalore, leveraging Nischal’s contacts there – Klarity just hired a VP of Engineering to oversee the team. For the US, they are looking to expand headcount with a few data scientists and salespeople.

Klarity raised money in an oversubscribed round following the MIT delta v Demo Day that has helped them pursue these growth goals, but are already beginning to think about raising their next round as well as where to establish their home office – with Boston, New York and San Francisco and London on their radar.

bottom of page