THE AI DEPOT
Executive Summary: As a precursor to building out an omnichannel practice and the AI that directs it, there is a need for discovery and machine learning ready data. We call this the AI Depot.
The last two years have accelerated a shift in pharma away from placing the primary emphasis for Health Care Professional (HCP) engagement on the sales force. The sales rep role is not going away altogether, but it will be coordinated with other channels to create a model where the focus is placed on individual HCP channel preferences and message interests. There is a human element to this change that we cannot forget: we personally know sales reps whose jobs have changed or have been forced to find new careers because commercial pharma organizations are becoming more efficient and deliberate in how they promote their therapies.
What used to involve quarterly alignment adjustments and call plans will be replaced by frequently updated HCP policies and orchestrated “next best actions” across channels – all under the banner of omnichannel engagement.
Technology continues to drive innovation and efficiencies and pharma cannot attain omnichannel excellence without an effective AI strategy. The fundamental constraint on using AI is having the right data. Organizations typically rely on their IT department to provision data, but the tendency will be that the data for AI will not be available in the required form in either the data lake or the data warehouse.
So, what is the difference between a data warehouse and a data lake? Here’s a quick recap:
Data Warehouse (DW) - highly structured, inflexible, used for standard reporting and analytics
Data Lake (DL) - minimal structure, flexible, used to capture “all” the data for further refinement in other repositories and applications (i.e. the data warehouse)
While these forms of data storage and organization have been the mainstays for years, neither is suited for AI. If the data scientist tries to go to the data warehouse, much of the data will be missing, thrown away because it was not required for aggregated reporting. If the data scientist goes to the data lake, all the data will be there (theoretically) but they will spend 80% of their development time trying to understand and integrate the data for the AI purpose at hand…and they will get it wrong. Also, it will be a one-off effort repeated many times across many data scientists.
There is a requirement for an AI-focused environment that borrows elements from the Data Lake and Data Warehouse and incorporates new ones – let’s call it the AI Depot:
AI Depot - semi-structured, flexible, used to rapidly incorporate and integrate new and updated data sets for discovery and AI delivery
The primary characteristics of the AI Depot are:
Rapid ingestion and provisioning of new data sources that may not even exist in the data lake
Breadth - every data element could be important to AI and needs to be explored, or at least understood
The lowest level of granularity possible
Frequently updated
Consolidated customer attributes - every characteristic about the customer is collected in one place
Reusable discovery - Tables can be organized around an area of analysis; there is one version of the starting point for all data scientists
Ability to rapidly generate data in required AI structure - A uniformly structured table with daily interactions for the last 3 years for each HCP
The data undergoes extensive QA for AI usability
Extensive use of history for AI results and versioning means that model results can be compared
The promotional approaches being used in pharma marketing are changing, and the data management approaches needed to support advanced, AI-informed insights need to change as well. Organizations that cling to traditional practices will not be able to implement omnichannel solutions if the data streams and structures do not exist. An AI depot is the key to starting the omnichannel journey.
----
Trusted by life science companies, Sentier provides a data-driven understanding of their customers. For the past 5 years, we have been delivering disruptive insights through innovative AI-powered solutions. Always at the forefront of the most effective data integration and machine learning approaches, we have created unrivaled models. As a result, our clients are better able to optimize their marketing and sales resources, and build stronger customer relationships.
Sentier is committed to the concept of High-Velocity Decision Making and we believe it can only become a reality with a strong and innovative analytics component.
We strive to be the leaders in the actionable application of new and emerging data and analytics approaches and to remove the barriers to our clients of benefitting from these solutions.
We believe that the ethical application of our services will benefit patients and doctors as well as the pharma and biotech companies we serve.
Tagged: data analytics, Data Lake, Data Warehouse, AI Depot, AI
