Select Page

Complexity Series: Abnormality as the Norm (pt.4)

Jul 2, 2021 | Blog

Having identified the challenges and opportunities provided by complexity, we can provide more detail on how IntelleXt generates solutions to the former while leveraging the latter. Core to that premise is the application of bleeding-edge calculation and modeling techniques to find solutions hidden from human judgement. While each application area deserves special attention, what they share in common are behavioral modeling and deep learning. 

Looking in the Mirror 

Macro is complicated; micro is simpler. This mantra cleanly summarizes our approach to behavioral modeling. The large-scale questions such as “should I sign with this company” are macro questions; too complicated for algorithms, they are the primary reason businesses have negotiators to begin with. For smaller-scale questions such as “what is the best price I can get under such-and-such conditions”, though, there exist micro processes to answer them which involve a smaller number of variables. This allows systems to model parties within a finite space, and use deep learning to compute optimal recommendations based on your values

Think of a sophisticated process, like driving a car. There are numerous different sub-processes involved: managing speed and direction to stay on the road are obvious mechanical ones, but other cars, pedestrians, signs etc act as inputs that must be considered as well. Each of the above aspects is a micro process, which could be calculated independently and merged together to generate a combined solution path. We already have autonomous cars in many places, demonstrating the technological ability in adapting to each of those inputs. But a self-driving car still can’t tell you where to go! Behavioral modeling is much the same; each component can be captured and combined in our systems, but you’re still in the driver’s seat. Once you decide on the what, we can then calculate the best how. And the method we use to determine that largely stems from innovations in deep learning.

Rocks That Can Think

Deep learning is a computation technology which can generate uniquely powerful insights into any application area. (For more information about deep learning, you can read this blog post.) From chess to the stock market, this technique has allowed computers to master some challenges more effectively than even the highest performance quartile of humans. For each question type, deep learning can be paired with specific datasets and calculation techniques to resolve those individual micro problems.

In the context of quantitative problems such as price, it’s possible to use behavioral modeling architecture to convert these problems into simulated scenarios. We can then discover the common patterns among the most efficient solutions and use them to generate novel scenarios. These new options optimize the balance between your and your counterparty’s values, assuring that every negotiator is satisfied with the outcome. 

For qualitative problems such as establishing a counterparty’s personality, deep learning can “read into” people using principles from the social sciences. Conversation between parties can be analyzed by engines to determine what typical features are present in the person you’re dealing with. When paired with insights from psychology and communication theory, this data can help make recommendations for what you should say or do in a given situation once we know what your goals are. Transparency and tact are both key to effective communication in negotiation, and our systems can help you get there. 

More broadly, deep learning enables our engines to extract insights that are embedded in data networks too complicated to be understood in a realistic time frame using traditional methods. This allows negotiators to commit all their resources to the (arguably more challenging) tasks of making long term plans and developing trusting relationships. All the while letting machines organize the information needed to make those tough calls and assist in developing an action plan once those calls have been made. 

Wisdom in Action

From both a business-to-business angle as well as between stakeholders, computer-based modeling and learning can reduce the complexity of negotiation down to micro-problems and provide situation specific solutions. All inputs included, opinions weighed, and outcomes calculated. Whether handling interpersonal dynamics, calculating trade-offs to maximize profitability, or trying to judge whether a short or long-term arrangement is preferable, we have you covered. 

Complexity is also controlled via our streamlined user flow. Other platforms only assist in one or two of the areas we augment, which results in the necessity to supplement those services with other means. This solution, while better than the default, still causes issues with communication breakdowns, information loss, and the frustration of having even more data to juggle. IntelleXt provides our unique deep learning solutions in tandem with contract management, message consolidation, video calling, and specialized educational resources for a clean, efficient workflow. Negotiation is never easy, but it doesn’t have to be confusing and stressful. With the aid of technology, our platform can supercharge you and your business with decades of insight in ways that are both accessible and intuitive. 

About Intellext™

Intellext is an AI startup that is revolutionizing the way contracts are negotiated, accelerating time to close, and improving deal terms. Intellext’s Intelligent Negotiation Platform™ eliminates the complexities of contract redlines and stakeholder collaboration and optimizes deal terms by applying machine learning during the negotiation process.

Joseph Hunt
Written by Joseph Hunt

Behavioral Researcher with formal training in psychology, philosophy, and user experience. Passionate about modeling behavioral and decision-making processes in pragmatic, actionable ways. BA in Psychology, San Jose State University

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *