Cognigy's Innovation Pillars: Powering Our Research-Driven Product Strategy

Innovation is at the core of a company's competitiveness. Driven by R&D it is a process that determines how the company searches for novel problems and solutions, synthesizes new ideas into a business concept, and eventually crafts them into successful products.

At Cognigy, our R&D mission is simple: Provide the most advanced, most flexible Customer Service Automation Platform to help global enterprises service their customers in a fast, frictionless & automated way. Of course, we leverage our experience and expertise within the Conversational AI space. But our mission and ambition drive us to focus on long-term goals. We believe that an AI-driven future deserves a sizable investment. This is the reason our business devotes a very significant portion of total earnings to R&D activities aimed at discovering and developing new AI capabilities and research initiatives. These state-of-the-art innovations and research initiatives are led by our R&D team who work relentlessly on ground-breaking technologies, making us a leader in Conversational Automation

In this article we will outline Cognigy's approach to innovation and explain our four pillars for innovation: Core Technology Management, Academic Community, Partner & Customer Network and AI Research.

Core Technology Management: Breaking New Ground with Patented Innovations

Since our founding, Cognigy has amassed unrivaled expertise in automating customer and employee conversations, all of which are backed by various Conversational AI patents. A prime example is the Cognigy Intent Analyzer. One of our advanced capabilities provides predictive insights into the NLU model performance. It is used to optimize user input recognition even before deploying to production and ensures maximum NLU performance with minimal training effort and keeps even the most complex enterprise conversational projects on track to success.

Another patented outcome of our in-depth research efforts is “Reconfirmation Questions and self-learning sentences”. This integrated functionality bolsters our platform by employing statistical relevance evaluations to enhance intent identification over numerous established user encounters and storing them for subsequent use. Essentially, it enables the AI to get smarter over time, fully autonomously.

Collaboration with the Academic Community

Cognigy lays utmost importance on knowledge transfer with scientific institutes. It encourages an open and collaborative research environment where employees collaborate with external partners and universities. This has led Cognigy to engage in a number of cooperations, to mention two exciting projects:

• University of Applied Sciences Duesseldorf: DISTEL - Data-driven, intelligent Storytelling with Robots
  DISTEL is a research project sponsored by the European Union that aims at using innovative AI technologies to tell multimedia stories that can be experienced emphatically by employing immersive VR/AR technologies and interactive robotics.
  
  
• Kassel University: HISS - Hybrid Intelligence Service Support
  The aim of this project (funded by the German Federal Ministry of Education and Research) is to develop and test innovative services and business models. In the HISS project, a concept is being developed that combines adaptive artificial intelligence (AI) via state-of-the-art bot technology with classic support.

Leveraging our Network of Partners and Customers 

Just as important, Cognigy's vast, resourceful, and varied network of partners like Accenture, Deloitte, and Deutsche Telekom are vital to innovation. Pioneering use cases developed with implementation partners & customers are crucial to in-depth research projects and successful products.

We also conduct product feature research in collaboration with our customers which entails empathizing and analyzing customer needs, developing features to meet those needs, and putting potential product features to the test with customers. In this way, we are able to better grasp the benefits that translate into genuine value for the customer. The approach of close cooperation with customers has, for example, helped us craft our conversational AI analytics stack - Cognigy Insights.

Cognigy’s AI Research Process

Investing time and money on R&D does not ensure success. Extensive market research to discover customer needs and aspirations are the keys to effective R&D. As customer preferences always change, we evaluate research on a regular basis.

The research method at Cognigy is user-centric, with customer pain points, feedbacks, and needs forming an essential part of the research. Cognigy also spends significantly in user analytics, data science, and market research to come up with research subjects that properly pinpoint our customer requirements and desires.

From research through delivery, new features or products go through a tiered process before being released to the market. Rapid prototyping, design thinking, and an open, collaborative atmosphere are encouraged during the discovery or research stage, which frequently involves collaboration with universities.

In the area of machine learning, research initiatives strive to create a virtuous cycle of iteration. Starting with the formulation of a question, problem, or idea, then developing appropriate measurement metrics and targets, and lastly iterating on algorithms and product design quickly.

Whenever we can we benchmark our research results and always strive to raise the performance bar. We tested our state-of-the-art multilingual NLU with market leaders like Microsoft LUIS, Google Dialogflow, and IBM Watson. The end results depicted that Cognigy's recognition of the intent was far more accurate, and the confidence was constantly high. 

In addition, our innovative conversational AI technology stands in line with trusted values and ethics. AI safety, explainability, and privacy guide our actions. As a result, Cognigy.AI fulfilled the 7 key criteria to gain a Trustworthy AI status set by the AI Cloud Service Compliance Criteria Catalogue (AIC4) audited by PricewaterhouseCoopers (PwC).

How We Transform Research Findings into Product Features 

During the discovery phase, we compile all the results from our extensive research into specifications or design documents, which are frequently accompanied by a prototype or proof of concept. These requirements are subsequently subjected to peer review open to the whole AI and engineering team.  

When the specification is approved, it is entered into an agile, customer-centric delivery process in which engineers work in biweekly sprints. Features go through full development, quality assurance, and deployment lifecycle where engineers are responsible for running their own code in collaboration with site reliability engineering and support teams. Upon conducting all these crucial steps, products are then ready for the market launch. 

We Embrace Innovation

Cognigy is committed to innovation and believes that allows us to create more efficient products and new ways of delivering our services. Our innovation excellence is the result of a multi-year effort involving most, if not all, aspects of the company. As a result, our products are backed by extensive research, allowing us to provide superior customer experiences.