IBM Research-Ireland

Building a winning team using AI

Share this post:


Figure 1: Opportunity Team Builder Interface

Building a successful team isn’t just about finding experts, it is about building a team made up of people who fill different but compatible roles. Team members bring with them their diverse experience, knowledge and social connections which benefit from the team, but also people have to have a certain level of overlapping knowledge and perspectives to be able to work together effectively. If you bring together a group of experts from different fields but their opinions are so far apart they can’t even talk about the problem because of different terminology or view points then it is difficult for them to find common ground. On the flip side if you bring together a group of experts who all agree with each other and have the same skills and knowledge then diversity of opinion and the ability to tackle unseen challenges can impact creativity.

Inspired by work on team composition by researchers such as Professor Sinan Aral , MIT Sloan School of Management, our Cognitive Analytics team at IBM Research – Ireland are are looking at ways AI can be used to recommend team members taking into account diverse experience, knowledge and social connections.  My colleagues Oznur Alkan, Inge Vejsbjerg, and I are working with IBM’s internal sales organization to build winning sales teams that can turn potential opportunities into winning deals.

Forming the right sales team for a new opportunity is vital and depends on understanding the roles required for the opportunity and then finding the right people to fill those roles such as managing the relationship with the client, having a deep knowledge of the product or being able provide an overall technical architecture seeing how all the products can fit and work together. Our team have developed an Opportunity Team Builder (Figure 1) using AI to support sellers in identifying required roles for the opportunity based on the products that the client is interested in, recommending the best people to fulfil these roles, and predicting a win probability based on the current team composition to guide users in team formation.

In a recently published paper called “Opportunity Team Builder for Sales Teams” presented at the ACM International Conference on Intelligent User Interfaces we describe how our AI tool analyses sales data to recommend a team that will maximize the probability of win. By using real data gathered from the sales division, we have built three models (Role Recommender, Team Recommender and Win Prediction Model) to work together within the Opportunity Team Builder. It not only recommends the best person to join a team, taking into account a combination of inferred skills and social relationships, but also the predicted impact the person can have on the overall performance of the team.

Opportunity Team Builder provides an intelligent user interface allowing the user to be a part of the interactive process when building the team. The user assesses the recommended roles and adds team members one at a time. When each team member is added, the expertise they bring to the team is taken into account to determine what gaps are remaining and need to be filled along with the social network of the team members.

The goal is to find the best expert to complement the current skills of the team while bringing together people from overlapping social networks. As a result, the user’s decisions become a part of the team formation process which not only increases user satisfaction but also increases the winning potential of the team.

The Opportunity Team Builder divides the team formation process into three main subtasks (Figure 2):

  • Determine which team roles are required to most successfully win a sales opportunity
  • Recommend a list of experts based on their skill set, social relationships and experience they have to fulfil the role
  • Predict the impact that adding a given person to the opportunity will have on the potential success of the opportunity

Figure 2: Assessing Impact of the Skills a New Team Member Brings to the Team

We deployed our solution on top of a real-world system and enabled it to be interactively used by the sellers. In order to evaluate our complete team recommender solution, we selected 2000 real sales opportunities and found that our system was able to assist users, new to their role, in finding experts to join their team without seeking recommendations and assistance from others. To validate this study we interviewed participants ranging in experience from sellers with only 5 months experience to over 30 years.

New sellers said that before using the tool they struggled with finding the right people to select for the team and would usually ask more experienced sellers for advice.  Sellers with established social networks in their field found the tool would help them in selecting people outside their network from different countries or divisions of the company. Overall, participants were impressed by having the ability to include personal preferences, institutional knowledge and being directly involved in the team building process. For instance, a seller may find that in their team selections previous relationships with the client were just as important as ensuring that the salesperson was a technical expert in that particular product.

Although the proposed AI solution is developed in the sales domain where the aim is to recommend seller-teams, the presented approach can easily be extended to support other domains such as recommending co-authors for research papers or building multi-disciplinary teams in the context of Care Co-ordination.



Oznur Alkan, Elizabeth Daly, Inge Vejsbjerg, Opportunity Team Builder for Sales Teams, Proceedings of the 2018 international conference on Intelligent User Interfaces. ACM, March 2018.












More IBM Research-Ireland stories

MIT-IBM Watson AI Lab Welcomes Inaugural Members

Two years in, and the MIT-IBM Watson AI Lab is now engaging with leading companies to advance AI research. Today, the Lab announced its new Membership Program with Boston Scientific, Nexplore, Refinitiv and Samsung as the first companies to join.

Continue reading

Adversarial Robustness 360 Toolbox v1.0: A Milestone in AI Security

IBM researchers published the first major release of the Adversarial Robustness 360 Toolbox (ART). Initially released in April 2018, ART is an open-source library for adversarial machine learning that provides researchers and developers with state-of-the-art tools to defend and verify AI models against adversarial attacks. ART addresses growing concerns about people’s trust in AI, specifically the security of AI in mission-critical applications.

Continue reading

Dark Matter Matters: AI Makes DNA Dark Matter Useful

What is the minimal description that captures a space? Asking a mathematician’s basic question of a  biological dataset reveals interesting answers about biology itself. This summarizes our underlying approach to subtyping hematological cancer. Disease subtyping is a central tenet of precision medicine, and is the challenging task of identifying and classifying patients with similar presentations […]

Continue reading