The new wave of generative AI products is already revolutionizing many aspects of our work, but a new kind of leadership is needed inside organizations to further guide this transformation.
During our startup journey at Hal9, we’ve had the privilege of sharing the front seat of this revolution with numerous Chief Data Officers (CDOs) who play a visionary role in shaping the future of businesses and the place of generative AI in them. CDOs have a very particular place at the intersection of data, operations, and future business vision. They were instrumental when we started shipping our enterprise data analytics product. Since the beginning, we realized that there were some constant themes in our conversations with them.
First, the generalized positive sentiment about the unrealized power of LLMs. Many CEOs who tried ChatGPT were asking: “Can we do this with our data?”, “How can we leverage this?”. Some CDOs’ first reaction was to consider building their own systems to harness this technology. Earlier this year, with the whole conversation about “wrapper” startups, it seemed like a couple of engineers could set up API calls and be done in a matter of weeks. These teams realized that a lot was missing around LLMs to be built, mostly when it comes to using them in the enterprise’s modern cloud stack with enough security, visibility, and resilience. They concluded that embarking on such an endeavor was not a smart use of resources. Not even mentioning figuring out how to solve hallucinations.
Secondly, plenty of businesses – regardless of their size or age – have gone through a data maturity process and now face a problem in properly digesting and leveraging their data. They moved on from spreadsheets to data lakes, data warehouses, visualization tools, etc. Yet, the challenge remains in maximizing the potential of these resources to derive strategic value and actionable insights without having to train everyone in advanced data tooling or fight for extra budget to grow the data team. Even with an infinite budget, we all know that requesting reports or new dashboards to the data team always takes time – for some companies this can take months. The data team, on the other hand, would rather focus their attention on more interesting and strategic projects than “urgent” requests from the business side of the company.
Fortunately, we had this knowledge to boost Hal9’s vision of reimagining enterprise analytics with generative AI. Our product empowers business users to ask questions to their enterprise database and get instantly generated dashboards. As you can imagine, the data security aspect was vital to our customers, so we developed technology that insulates their data and makes sure that nothing leaves their cloud while allowing them to leverage LLMs hosted by third parties like GPT4 or LLaMa. We also solved the hallucination problem by making the LLM generate code instead of direct answers to the user, so Hal9 responses are always accurate and based on auditable code.
This aligns with our expectation about a future of diverse LLMs that will become a new resource for the company, similar to cloud computing. We’ve realized that CDOs will become the orchestrators of multiple models and tools used for different purposes inside the organization, managing their data access settings, permissions, and budgets. Just as they are using Hal9 for their data analytics, they might use other tools for content generation, marketing copywriting, code development, or legal document editing.
Another lesson we’ve learned is that this AI revolution is very industry-agnostic. Every company with enough data maturity can leverage Hal9, often with greater results in established segments such as legal or insurance. For example, Coterie Insurance’s CDO Paul Bessire and Vividly’s Co-Founder Alex Whatley, are some of the early adopters of this technology and state that:
“By leveraging generative AI, we significantly enhance our ability to make swift, effective, and cost-effective decisions, maximizing the quality of service we offer to our agents, policyholders, and carrier partners.” — Paul Bessire CDO at Coterie Insurance (Small business insurance, simplified)
“AI was already central to our product, generative AI will supercharge the ROI we deliver to our customers” — Alex Whatley Co-Founder at Vividly (Trade promotion management)
It is very exciting to imagine what will come next. We at Hal9 are envisioning moving forward from answering descriptive questions (i.e. How did our sales perform in x region last month?) to predictive questions (How will our sales perform next month?) and eventually prescriptive questions (How can we improve our sales performance?). Meanwhile, we are sure that the role of the CDOs will continue to be central to the generative enterprise revolution, making us wonder when (not if) we’ll meet the first Chief GenAI Officer.
