ayfie: Unleashing the Potential of Linguistic Integrity

Robert Wescott, CRO

“Language is the currency of law and linguistic integrity is crucial in legal search and analytics,” says Robert Wescott, the CRO of ayfie, a New York-based legal technology firm. Current commonly used technology in eDiscovery–LSI and its Bag-of-words (BoW) approach–have severe limitations regarding due their lack of natural logic of grammar and semantic meaning. With its unique linguistic foundation, ayfie Inspector —a next-generation text analytics and machine learning platform—is disrupting the status quo in eDiscovery. Backed by more than 30 years of applied research in computational linguistics at renowned universities, ayfie’s analytics are built on a deep understanding of the underlying language patterns inside the content. “ayfie Inspector layers machine learning onto a rich understanding of what is meaningful,” Wescott adds. This enriched linguistic foundation has many advantages over statistical systems because it is able to surface more precise and meaningful concepts and connections.

ayfie provides superior information retrieval fidelity in a portable platform. According to Wescott, “It’s ayfie’s goal to be ubiquitous in eDiscovery, compliance, and information goverance: as a stand alone information explorer, OEM partnerships with software companies,and interfaces to leading applications, such as Relativity.”

Technology Assisted Review (TAR) is a supervised machine learning-approach for iteratively identifying responsive documents. ayfie’s classification tools combine proven algorithms (logistic regression and support vector machines) with a precise semantic structure, to enable efficient TAR workflows.

Unlike the BoW approaches like LSI, ayfie layers machine learning on a richer semantic understanding of natural language

Combining this with other key eDiscovery components, ayfie Inspector offers a system that unifies email threading, entity extraction and normalization, clustering, conceptual search, near-duplicate detection, classification (TAR), query/keyword suggest, and faceted search.

Indicating towards industry demands, Wescott points out, “Investigators want evidence in and out of the data quickly; they demand swift gains from relevant information, save those results, and batch them for further review.” Perceiving this, ayfie Inspector delivers exceptional structure to its text search functionality, which in return, can rapidly surface meaningful findings. Extensions to the platform include contract analysis and PII detection. In addition, ayfie’s platform is highly performant with scalable storage and processing technologies that cope with massive datasets.

The company keeps its engineering and product development processes agile so that they can respond to the clients’ evolving workflows in eDiscovery for maximum impact. For instance, ayfie provided its advanced research toolset to enhance the innovative product portfolio of The Economist Group —CQ Roll Call. ayfie Inspector powers CQ’s ingestion, entity extraction, normalization, and classification capabilities and will extend this to include search and alerting. In another scenario, ayfie partnered with One Discovery to strengthen the organizations' ECA (Early Case Assessment) and TAR (Technology Aided Review) processes and deliver unprecedented visibility into the responsiveness of documents. Furthermore, ayfie Inspector for eDiscovery surfaces critical content intelligence directly within One Discovery for better transparency, control, and precision.

“We tend to partner with our customers to deeply comprehend their challenges and work together to respond to their needs dedicatedly,” asserts Wescott as he stresses on ayfie’s unique propositions in the legal market. Today, ayfie contemplates the possibilities of leveraging its platform to solve challenges with unstructured data in other industries, turning ayfie’s unique approach to extracting structure from text to compliance use cases in the healthcare, insurance, and financial industries.