Technology

The Soffos platform provides a low-code solution for building advanced NLP features into applications. NLP stands for Natural Language Processing and refers to AI that has to do with human language as it appears in text, video, and audio formats, whether within documents or online.

Our NLP toolkit consists of various modules which are available to developers as RESTFUL APIs. Developers can plug into these APIs to add features as outlined on our modules page. For ideas on the types of solutions that can be built using our modules, we recommend visiting our use cases page.

Our goal is for our developer dashboard to be the easiest to use, most comprehensive and most powerful NLP toolkit on the market. Developers who have found other APIs on the market lacking in either performance or variety will find everything they need on Soffos’ platform for the solutions they’d like to build. Our technology creates content from resources that are not specially prepared for use with virtually no human intervention other than uploading or copy and pasting from existing documents. The only requirements are that the resources be readable and not just in the form of data, images, charts, or slides with a handful of bullet points.

Our computational linguistic strategies then use distributional semantics to generate, answer, and score questions using novel symbolic computing methods to clean up messy machine learning (ML) output using fully proprietary logic frameworksDistributional semantics is a theory of meaning that is computationally implementable and exceptionally good at modeling what humans do when making similarity judgments regarding words or phrases.
‍
Our work focuses on syntax theory, the intersection of compositional and distributional semantics, logic programming, and automated inference.

While we leverage standard ML techniques (transformer encoder/decoders, vector-space word embeddings, topic clustering, etc.), we are also developing error correction frameworks that target the syntax-semantics interface where large language models usually fail.

We plan to expand the techniques we are developing to incorporate taxonomies and ontologies that our customers bring to the table for tasks such as Named-entity recognition (NER) and document classification.

We have chosen to standardize on processing frameworks that use the most logically and mathematically pure theories of syntax available, to ensure clean translation of natural language into logical relations.

The team at Soffos are all about finding solutions for your business. Our application architects create these applications using our modular API framework to engage the Soffos’ platform where our AI listens, reads, processes, questions, understands, and converses on any subject. Our platform is accessible and consistent regardless of the underlying technology. Modules currently in development establish a standardized low-code interface and may initially leverage 3^rd party algorithms while we develop advanced replacements to seamlessly upgrade the resulting experience without impacting the overarching application. This means your application gets to market earlier and enables it to stay ahead of the competition.

We empower your users to query, access, and apply knowledge stored in unstructured text, so they don’t have to rely on keyword search or statistical text-mining tools.

The result is the ability of AI to present information in natural language, using theSocratic Method to ensure that information is fully understood and retained by the user.

Given that few NLP companies are exploring areas such as logic programming and formal semantics, we are securing patents and publishing both academic papers and technical white papers on our methods.
‍
We have filed two patents that are pending with the USPTO: