Skip to Content

A   Differentiated   Approach to AI Based on Decades of Experience

DraftSpark RFP Assistant is the first in a family of Collaborative Authoring tools from Xerox, designed to assist knowledge workers in drafting high-value business documents.

image of 3 persons understanding a differentiated approach to AI

Deep Expertise in Document Modeling and Natural Language Processing

image of sketched city

Scientists at the Palo Alto Research Center (PARC) have been developing AI since the 1970s and began coding the very first Natural Language Processing (NLP) software soon after. The spark for this foundational work was an understanding of how people might use digital capabilities to support them in working better.

The scientists' goal then – and now – is to understand how knowledge workers create documents, then enable machines to help them do it more effectively. Computers can perform the repetitive, mundane tasks, like parsing the document, perusing the archive of prior proposals for the best content, detailing the RFP submission schedule, and creating content blocks and tasks. Knowledge workers are then empowered to focus on what they do best, creatively developing the best descriptions of their organization's capabilities.

DraftSpark is just the first in a family of products coming from Xerox that follows this paradigm of human-machine collaboration.

sketch of human machine collaboration

Human-Machine Collaboration

We believe that AI can augment people and help them do more complex work more effectively. Collaborative Authoring is a key area where we are working to enable knowledge workers to focus on higher values tasks in creating critical business documents.

sketch of hybrid AI

Hybrid AI

An approach to AI that combines knowledge-based models with machine learning. This technique is essential to applying AI to situations where there is limited data, making machine learning alone ineffective.

sketch of interactive machine learning

Interactive Machine Learning

PARC’s interactive learning systems leverage domain understanding and contextual information to make machine learning more efficient. This enables effective training of models with fewer examples, enabling accurate results from smaller data sets.

Do NOT follow this link or you will be banned from the site!