For many giant enterprises, AI transformation is a narrative of silos — enterprise models working competing experiments, knowledge sitting in disconnected programs, and outcomes staying out of attain.
Cushman & Wakefield has taken a distinct path. Sal Companieh, Chief Digital and Data Officer for the worldwide industrial actual property providers agency, has spent the previous 4 years constructing an enterprise AI core designed for reliable, sturdy, and scalable affect throughout 53,000 colleagues worldwide.
Sal sat down with CIO.com to debate how the core is constructed on a deliberate working mannequin, unified knowledge technique, and a Databricks partnership that goes far past options.
CIO.com: How did you transition from fragmented AI efforts to a centralized enterprise AI core, and what had been the most important challenges?
Sal Companieh: After I took this function 4 years in the past, I instituted a product working mannequin that embedded technologists in each enterprise unit. The targets had been to rebuild connectivity, belief, and business-forward pondering on each the info and experiential sides, and to offer us the flexibility to co-create. I ensured that technologists had accountability for income and EBITDA; they introduced the creativity, the ingenuity, the agitation.
Whereas many organizations had been going by way of an “AI pilot craze,” we deliberately maintained a top-down focus, which helped us proceed to garner belief whereas additionally augmenting our knowledge. Most firms had been working pilots. We had been constructing the inspiration that will make each pilot price one thing.
When AI adoption surged, we had been deliberately top-down — what we name the “Cushman Method.” We recognized the biggest go-to-market or worker expertise transformations and attacked them first, serving to us construct belief and strengthen our knowledge basis. The largest problem was managing the maturity variations throughout the group. Whereas everybody targeted on expertise, we had been anchored on human habits and producing belief.
CIO.com: What macro developments are shaping Cushman & Wakefield’s digital and AI methods immediately?
Sal: Determination-making on the investor aspect is maturing and changing into extremely data-centric. Throughout providers organizations, we’re seeing a pivot from relationship-based shopping for to intelligence- and insights-based shopping for. Our potential to distinguish by way of partnership and authenticity stays a continuing consumer demand.
For us, the AI surge taking place out there was a tempo accelerator and never a pivot in technique. We now have maintained the identical technique and working mannequin from the beginning; it’s simply that now the expertise has lastly caught as much as our ambition. That distinction issues as a result of it means we haven’t been chasing a pattern. We now have been executing on a plan we’d already dedicated to.
About three and a half years in the past, we shifted how we approached capital funding for expertise. Exterior of cyber and infrastructure, all the things needed to be co-created and co-presented with a enterprise chief. That stored us totally aligned with enterprise priorities. There may be one enterprise set of precedence outcomes for the corporate; expertise is one part of delivering on it. Each technologist can draw a direct line from what we are saying on earnings calls to the work they do on daily basis.
CIO.com: What methods have you ever applied to construct shared requirements and a typical platform whereas nonetheless enabling business-unit flexibility?
Sal Companieh: Three issues drive our method:
- the working mannequin, which prevents duplication and separation from the widespread platform;
- the monetary funding mannequin, which aligns capital to construct firmwide capabilities no matter which P&L advantages;
- enterprise requirements for tactics of working, the place structure groups guarantee expertise is approached as uniformly as attainable, with flexibility on the business-unit degree.
We’ve efficiently matured our working mannequin thrice. Flexibility itself is a talent we’re constantly constructing.
CIO.com: How does Databricks match into your technique?
Sal: Our potential to construct Lego bricks of various capabilities and match them collectively for each enterprise unit was our foundational technique. On this means, Databricks suits into three areas. First, I imagine the phrase “companion” ought to imply one thing, and when it comes to our partnership with Databricks, it means management, tradition, and the flexibility to genuinely co-create with us, in addition to their understanding of our maturity degree and the tempo at which we wish to remodel.
Second, we had been impressed with their product roadmap. We weren’t shopping for only a product suite for immediately’s wants; we wished to align their funding technique to the capabilities we would have liked to maneuver ahead. Third, precise function performance. Our intelligence layer is vital to our progress and differentiation.
Databricks lets us pull completely different levers relying on the enterprise unit we’re supporting. We’re utilizing Genie to place trusted knowledge straight into the palms of the enterprise by simplifying knowledge high quality and knowledge governance workflows. Workers can use pure language queries to establish lacking or inconsistent information, validate knowledge high quality throughout programs, evaluate governance insurance policies, and monitor compliance metrics while not having deep technical experience. This allows enterprise customers to rapidly discover advanced datasets, enhance confidence in enterprise knowledge, and drive sooner, extra knowledgeable choices.
CIO.com: How has the Databricks structure particularly formed the way in which you unify and govern knowledge at enterprise scale?
Sal: Our basic speculation is to tether our group’s data on a worldwide scale. We all know that is materially useful to ourselves and our shoppers’ wants, so we’d like one of the best platforms to guard and allow this data. That’s the place the Databricks platform is available in.
We’re a 108-year-old firm whose historical past would recommend knowledge strikes by way of a human provide chain throughout the group. Databricks has allowed us to digitize the motion of insights, piece by piece. Our intelligence layer is vital to our progress and differentiation. Being able to construct up capabilities and match them collectively in a different way for each enterprise unit — whereas retaining the widespread platform intact — is our foundational technique.
The keenness for AI hasn’t subsided, however it has been wellrounded with a rising recognition that wholesome, ruled, scalable knowledge is what really accelerates outcomes. Databricks is central to how we implement that uniformly throughout the group, and we are able to flex the place wanted on the business-unit degree.
CIO.com: Are you able to share examples of measurable outcomes from this transformation?
Sal: The time from thought to consequence has gone from months to days. We now have a gathering, ideate, and inside the following week, we’re delivering worth to the enterprise. For instance, the flexibility to onboard, combine, and activate new shoppers and acquisitions has materially lowered. However an important consequence is the shift in human habits. We’re not preventing the change continuum anymore. A query that traditionally would have required 5 telephone calls, three emails, and two Groups chats to reply is now at our leaders’ fingertips. It’s a continuing evolution. We’re ensuring there’s much less “organ rejection” to vary as a result of at this tempo, change has to turn into an inherent, on a regular basis exercise.
CIO.com: What recommendation would you give different IT leaders trying to construct and scale an enterprise AI core?
Sal: Don’t underestimate the human change; persons are partaking with a wide range of data baselines and fears. Genuinely educating folks on each the chance and the foundational work that’s required to seize it’s a non-negotiable. Our language must proceed to evolve and mature as we maintain extremely impactful roles for shaping the way in which work is completed and industries are formed for the following technology.
This can be a second in time the place all the provide chain is reworking concurrently: your suppliers, your shoppers, and your individual worker base. Sustaining an outside-in lens whereas balancing inside-out goes to be vital. Leaders who acknowledge that and construct for it’s going to outline what comes subsequent.
To find how greater than 25 trade consultants are charting a course towards profitable AI deployment, entry the “Making AI Ship” report from Economist Enterprise, produced with help from Databricks.