Ontra Synapse AI

Ontra Synapse is included in Contract Automation and Insight at no additional fee. We’re currently embedding the technology in FundFormer as well and will be making additional announcements about this product in coming months.

Ontra Synapse requires no setup to perform basic contract analysis since models have been pretrained on anonymized and aggregated data from Ontra’s industry leading contact repository. If customers wish to access more advanced functionality such as automated precedent search, they will first need to upload previously negotiated contracts into Ontra’s contract repository. Ontra Synapse will analyze these documents and incorporate associated data into future responses.

Ontra began to integrate AI technologies into its products in 2016. During the ensuing years, we've built a talented team of machine learning engineers. These professionals have joined forces with Ontra's deep bench of former lawyers, i-bankers, and asset managers to pioneer innovative new ways to apply machine learning to the legal domain. We've incorporated AI into our solutions and continue to invest heavily in the engineering talent and associated research required to push the boundaries in this space.

Nuanced legal text: Nuances in legal language such as esoteric terminology and variations in clause construction present a unique challenge for AI. Machine learning models that are trained on non-legal data sets are likely unable to consistently identify and understand legal clauses. The unique data used to train Ontra models allows Ontra Synapse to overcome this challenge. Ontra’s models are trained on aggregated and anonymized information derived from over 800,000 contract negotiations. The unprecedented breadth and depth of this dataset underpins the superior precision Ontra’s models deliver. Additionally, while legal contracts are highly structured, open source models typically don’t recognize these structures. Consequently, open source models fail to sufficiently address use cases such as document summarization where results must be correctly categorized and organized. To address this issue, Ontra has made an extensive investment in mapping “anatomies” for routine contracts. These anatomies are structured representations of key terms associated with routine contracts and generally contain hundreds of fields.

Industry-specific tasks: Legal workflows include many tasks that are unique to the legal profession such as marking up a document. While open source models address a range of generic use cases, they must be modified to answer industry-specific questions. Ontra is uniquely positioned to address the nuances associated with private markets legal workflows. The Ontra team includes former investment bankers, GCs, dealmakers, and lawyers. Our first-hand experience allows us to deliver tailored solutions with unrivaled depth and sophistication.

High precision: Perhaps most importantly, when it comes to contracts, there’s no margin for error. While “mostly correct” may suffice in other industries, that’s not the case when dealing with private funds contracts where the costs associated with legal errors can be massive. Ontra Synapse meets the exacting standards of the private markets via an innovative human-in-the-loop system.

Ontra Synapse differentiates itself from other AI solutions in two important ways. First, its proprietary models have been designed to generate answers specific to private markets legal workflows. Second, these models are trained on Ontra’s unique data set — the industry’s leading repository of aggregated and anonymized routine contract data. The combination of these two factors allows Ontra Synapse to generate outputs that outshine any other solution in terms of relevance and accuracy.