“If you can’t explain it simply, you don’t understand it well enough”—Albert Einstein.
Today, artificial intelligence (AI) solution providers are omnipresent, however, most of the solutions they provide fail to explain the ‘why’ and ‘how’ of the recommended decisions relative to the organizational goals. Especially, the CFOs of financial, advertising and real estate companies have to face countless challenges in having a robust AI strategy in place. The Investments in AI-based solutions are similar to those in R&D, because “One cannot assess the outcome in advance before trying,” says Dr. Foster Provost, a Founder of Detectica, Inc.
“The “knowledge” in AI systems is created by some combination of human judgment and machine learning from data, and thus all AI systems have blindspots, which arise from human biases and/or biases in data collection and processing. The first step in mitigating the risk from these “unknown unknowns” is to find them. “Detectica, with their unique suite of services and patented technology, helps organizations lift the shroud of uncertainty involving their AI investments.” “For example, where are our AI investments most likely to pay off quickly? The cleanest applications of AI to improve business outcomes are cases where the data necessary to make the decisions and the decision-making interface are already automated, and ideally in the same system.” Based on Detectica’s advice and guidance, organizations can either choose to create an AI solution on their own, have someone else build it for them it, or outsource it to Detectica.
According to Dr. Provost, AI solutions often involve not just the primary investments in building the system, but many complimentary investments: people, processes, and business transformation. Therefore, it’s imperative for a business to conduct a pilot project to assess the feasibility before implementing the solution. Detectica strongly advises CFOs to make smaller investments and do the necessary “homework” to mitigate the risk of failed investments in AI-led initiatives. Organizations can either choose to create an AI solution on their own, have someone else build it for them it, or outsource it to Detectica.
At the outset of a client engagement, Detectica works with the customers to plans a roadmap that may include building an in-house data science team. A key element of their engagement is the “stakeholder workshop” that guides clients on how stakeholders ought to think. The training program helps firms focus on opportunities, all while understanding the challenges. Not only does it facilitate the discussion of how AI/machine learning/data science may solve the problem, but also fill specific knowledge gaps on these technologies and work toward specific designs. “Many organizations already have a solid data strategy, but lack sufficient understanding of advanced analytics; Detectica helps them polish and implement their AI strategy,” says Dr. Provost.