Creating best in class Analytics product development teams

Teams exist to create a unified amalgamation of various skills

The reason teams exist is because it is supposed to bring people of varied skill sets together to achieve a common objective. In one way or the other, humans have being doing it for thousands of years.

However, our approach to creating teams, as per my perception, may be flawed. More so in this age of Digital explosion, when the type of solutions that teams need to create or the type of problems teams need to solve have changed significantly. We still look at what people have been doing in their roles (experience) vs what they are naturally wired to do. 

Time to bring more science to team building

So what does the wiring mentioned above mean ? The wiring in this context happens in the neurology of our brains and defines our areas of strengths. When we see that someone has drawn a beautiful portarit, we exclaim -“That person is so creative”. This creativity stems from brain wiring. A person can have several areas of strengths, which means that you can be creative AND few other things (Analytical, Interpersonal etc.).

If we have the data on this “wiring” for people and we can map it with their performance history in those areas, we will have invaluable resource to create “tailormade teams”.

Why the need to create these “tailored teams” ?

As the ecosystem around us evolves, the product lifecycles have decreased significantly. The Business challenges that entities face have increased significantly. Both these aspects mean that there is a need to be continuously on your heels and develop products and solutions iteratively. In summary – Business operations is becoming more and more Project based rather than day to day operations based.

As iterative project based work becomes the new normal, the importance  of putting together right teams has also increased significantly. Also, the natire of problems being tackled and/or solutions being developed has changed drastically as well, resulting in the need to bring more science into team building.

So how exactly do you capture this wiring and use it for team building ?

There are multiple frameworks that can help you capture the thinking processes/algnments of the brain. One such model is the HBDI model. As per Herrmann international definition, HBDI framework:

“The basis of Whole Brain® Thinking and all Herrmann International learning modules, the HBDI® teaches you how to communicate with those who think the same as you and those who think differently than you. Once an individual understands his or her thinking style preferences, the door is open to improved teamwork, leadership, customer relationships, creativity, problem solving, and other aspects of personal and interpersonal development.”

For our purposes, we will simplify the HBDI model, as shown in the figure below. The model believes that thinking process of the way the brain works, can be cateorized into four different quadrants. The strengths of each are shown (simplified). A person may have advanced capabilities in more than one quadrants and hence can be staffed during more than one phases.


The idea is that you have this assessment for each prospective team member, then you validate this with their historical work appraisals and feedback, and then chose the individuals that you need, based on their thinking strengths. Remember that:

You will need different strengths during different phases of your initiative.

Now let us use a Supply Chain example to see how this can be applied.

An applied example from a Supply Chain AI solution development  perspective

AI Product development lifecycle

As mentioned above, different phases of a project leverage different types of thinking. As per Larry Leifer, professor of design thinking at Stanford, typical activity phases of a Digital Product are as shown below.

  1. Shape the future
  2. Growth & Scale
  3. Operationalize & optimize
  4. Maximize value
  5. Evaluate the future


Since the skills needed during each phase, as shown in the diagram above, are diferent and hence varied level of skill sets need to be leveraged during each phase. Since no single person will probably have all these skills together, designing teams that fit tugether optimally like pieces of a puzzle can work its magic.

Project Objective

High Level team objective : What are the opportunities to use Reinforcement learning in Supply Chain analytics ?

Shape the future or Idea formulation:

The predominant thinking process needed is “Innovators”. This where the future is shaped with a vision. In this example, Betty has a vision for :

The Idea: The warehouse slotting system of the future in which an algorithm uses reinforcement learning, to train the warehouse robots in an automated warehouse, slots products to optimize space and productivity.


Betty focusses on the few functions of her idea initially, and maps the relevant actors in the ecosystem. As shown in the illustration above, Betty’s thinking process is primarily in the “Innovator” Quadrant and she has aptly leveraged it to think about an innovating application of reinforcement learning.

Growth and Scale:

Once an idea/concept is finalized as good candidate, proper socialization of the idea is needed. This can be achieved by “Humanitarians”. Socialization, Idea validation and selling the idea internally happens in this phase.

Validate and Sell the Algorithm logic : Initial blueprint of the logic is shared with floor automation engineers and warehouse operations managers for feasibility discussions. Once the logic, that will probably get iterated several times, gets a Green light from floor folks, it then gets shared up in the chain in the form of a business case, to get funding for a pilot.

Stacy, in tandem with Betty then takes the idea, do feasibility discussions and analysis with stakeholders, make the business case and gets the buy in. As seen in the illustration below, Stacy’s thinking process is a mix of humanitarian organizer, which is an apt fit for this stage of product development. However, note that the prominant thinking process for Stacy is Humanitarian, followed by organizer.


Operationalize and Optimize: 

This is a critical role where you need someone who has a good breadth across all the four quadrants, with good depth in “Organizers” and “Humanitarians”. 

Frame, Build, Pilot and interate: The product features that have been finalized in grw and scale phase now need to be finalized in the form of a framework. An initial prototype will be developed around this framework.

In this case, it will be the Algorithm developed using an open source deep reinforcement library like TensorFlow. There will be two aspects of project management process involved here. One is managing the code development process.

Stephanie is primarily  Organizer followed by Humanitarian which makes her a good fit for this phase. Being an organizer provides her great project management skills and being a humanitarian has given her good interpersonal skills needed in this phase.


Maximize value: 

This stage of the lifecycle is where after making sure that the tool is working the way it was supposed to, you evluate the output to draw business and operational insights  and how it will integrate with other processes and systems.

Output analysis, recommendations and implementation : The tool will recommended a set a rules for bot pickers, defined on aspects like velocity, form factor, product category, margin etc. of the SKUs. A theoretical feasibility analysis would have been done in the previous phase. This phase will focus on transforming efficiencies (like fewer sq. ft area, more throughput etc. in $ figure).

Effective planning also needs to be done to finalize how this will interface with interacting systems like WMS.

Clara is Primarily a thinker, that provides her the analytical expertise to evaluate the outputs and effectiveness of the developed tool and think about how the tool will integrate into the end to end Supply Chain planning processes, which will involve innovative aspects.


Evaluate future scenarios:

This phase is about how the product AND learnings and capabilities developed from this product developed can be leveraged across the wider enterprise.

Evaluate the path forward : What needs to be done now is : First, the team needs to understand what type of additional incremental developments can be made to this product in order to expand its capabilities within the DC environment. Next, you need to do an assessment how this new capability of designing operations AI solutions can be extended into other operations areas.

Samantha is a good fit to lead this since her primary thinking profile is Innovator, which is the primary need to extend the solution to the enterprise envoirnment. Het thinking aspect helps her formulate incremental development opportunities.



Views and perspectives my own.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s