Around the time of ChatGPTโ€™s release, Sanja Bogdanovic-Dinic, Head of Data Solutions at global product and digital engineering firm HTEC, noticed a change in client demands.ย โ€œPrior to ChatGPT, clients would come in and say, โ€˜We need to do this [create a product or feature] for our end customers.โ€™โ€ After ChatGPT, Sanja recalled a change. โ€œClients came in and said, โ€˜We need AI.โ€™โ€ She noticed this change happening across industriesโ€”HTEC works with clients in variousย fields, from LG to the World Health Organization, and startups in industries from retail to robotics. โ€œIn the clientsโ€™ minds, AI became this magic wand that you could just wave over anything, and it would magically work.โ€ Of course, Bogdanovic-Dinic continues, โ€œItโ€™s not quite like that.โ€ ย 

Since the launch of ChatGPT and other prominent AI technologies, much of the work done by HTEC has centered around educating clients about AI and, more importantly, showing them how to prepare and deploy enterprise AI efficiently and effectively. Part of that education has included releasing white papers, beginning with the piece titled โ€œData and AI Readiness Assessment: Gen AI is Here to Stayโ€” Is Your Data Ready?โ€ Here, Worth offers an exclusive look at the paper and insights from its author, Sanja Bogdanovic-Dinic.ย 

Bogdanovic-Dinic says itโ€™s essential to understand dataโ€™s key role in AI solutions from the start. โ€œPeople may think, โ€˜What does data have to do with it?โ€™ Theyโ€™ve heard that AI is built on data, but they also think that once youโ€™re using AI models, youโ€™re done dealing with the underlying data that feeds themโ€”far from it.โ€ Data will need to be continually cleansed, structured, and fed into AI-driven solutions, and the solutions require increasing amounts of unique data types. In the wake of the digital revolution, the quantity and variety of data has increasedโ€”especially with the addition of images and videos to traditional data types like text. 

Techonomy 24: Leading with Intelligence

ย 

To ensure oneโ€™s data is ready to create and maintain an AI solution, HTEC suggests clients evaluate how far along they are on the โ€œvalue generation journey,โ€ which is the process of positioning data as a value-driving asset. To determine this, the white paper proposes that organizations use a data maturity model (DMM), specifically one based on insights from Joe Reisโ€™ book Fundamentals of Data Engineering: Plan and Build Robust Data Systems.  

The DMM separates data maturity into three stages: 

  • Stage 1: Getting started: The first stage begins with newly collected data in its raw and unstructured form. The focus here is on cleansing the data (removing duplicate information and outdated details and verifying accuracy) and gathering it all in one place. Data reliability, hygiene, and trustworthiness are the goals.
  • Stage 2: Scaling with data: The second stage capitalizes on the cleaned-up data and uses it to build models and introduce basic analytics. Organizations can use the data in these models to understand things like monthly sales trends and operational performance. Businesses may also begin to make data-driven decisions based on their new AI capabilities. ย ย 
  • Stage 3: Leading with data: The third stage is the most advanced use of data and AI, where organizations are using sophisticated models for predictive analysis and proactive decision-making.ย 

Play Hard, Sleep Well: Burlington in Any Seasonย 

Vermontโ€™s largest city is known for ski weekends and foliage drives. A new lakefront hotel, serious kitchens, and a terroir-driven wine bar make it worth coming any time of year.

ย 

All of this sounds ideal, but how can companies get started?ย 

This is the subject of the paperโ€™s second half where HTEC presents its research-based โ€œData and AI Readiness Model.โ€ This model is broken into three levels of its own: โ€œFoundational,โ€ โ€œOperational,โ€ and โ€œTransformational.โ€ It diagnoses where a business is in terms of AI development and helps clients decide their data and AI objectives. Additionally, the white paper provides key examples to help readers troubleshoot issues at each stage of the model:  

  • Foundational level:ย At this early stage, companies may struggle with consistency in their data-gathering methods or accessibility.ย Here, HTEC provides recommendations for building a core infrastructure and foundational data platforms to collect and manage data. ย 
  • Operational level: In the second stage of the model, companies may struggle with extracting meaningful information from their data and converting it into actionable insights. HTEC recommends optimizing workflows with advanced analytics rooted in the latest data science and machine learning principles.ย 
  • Transformational level: When companies have reached the third level of the model, common challenges include leveraging data capabilities to identify growth opportunities. HTEC recommends building AI-powered platforms to explore novel solutions and considering business model shifts to support substantial offering changes.ย 

Wondering how this AI Readiness Model looks in action? The paper highlights an HTEC client success storyโ€”a maritime company that completed all three maturity model levels. HTEC began by helping the company set up a data integration process, which enabled them to digitize relevant processes and data. During the operational level, HTEC showed the company how to leverage real-time data insights to optimize voyage planning. In the final phase, HTEC developed a maritime management platform, empowering the client to launch a routing serviceโ€”a standalone product that the client could market to other organizations in the industry. 

Bogdanovic-Dinic emphasizes the importance of correctly using models like the one HTEC presents in the paper, as well as the importance of approaching the data and AI process with patience and a long-term perspective. โ€œWhen I’m talking to potential clients, they may ask, โ€˜Can we skip a step [in the data maturity model]?โ€™โ€ she says. โ€œBut if businesses donโ€™t assess data quality and other factors in the beginning or take the time to build basic models in the operational phase, theyโ€™re at risk of being tricked into the illusion of progress. Theyโ€™ll fall victim to a false sense of achievement where theyโ€™re only accessing the most surface data results, and that will result in missed opportunities.โ€ 

Get the full white paper at htec.com.ย 

Join Us

Sanja Bogdanovic-Dinic will be speaking at Techonomy 24: Leading with Intelligence at Civic Hall in NYC on November 19. Together with the Worth team we will touch on the most the significant issues of our time through deep and provocative discussions. One of the most pressing being AI.

Worth Article Data Readiness 4
Sanja Bogdanovic-Dinic

Techonomy 24 is a highly interactive, multidisciplinary event designed to equip leaders with the knowledge and insights they need to prepare for a better future. If you’d like to join us, please register here.