Here is California Business Journal's interview with Ghadi Hobeika, Managing Partner and CEO, Artefact US
Topic: Moving beyond “pilot purgatory” with agentic AI and a new operating model
Q: Welcome Ghadi. Thank you for sitting down with us today. Tell us a bit about yourself and about Artefact.
A: Artefact is a uniquely positioned and fast-growing AI-native consultancy that sits at the intersection of AI strategy and execution. Over the past decade, we’ve worked with over 1000 global clients across Facebook, LVMH, Carrefour, Hilton, Mars, L’Oreal, Nissan, Sanofi, Citibank & others. Since the public launch of LLMs and broader GenAI, our business has accelerated towards process reengineering and global business transformation through AI. As for me, after having started in Management Consulting and spent 12 years in retail, I’ve had the privilege of leading or North American business for the past 5 years.
Q: Let’s talk about Artefact’s latest announcement. Tell me about the Agentic AI Transformational Programs.
A: Artefact has designed and delivered Agentic AI Transformation Programs to help enterprises move beyond experimentation and deploy semi-autonomous, governed AI systems that operate inside core workflows. The timing is driven by what we’re seeing in the market: many organizations have proven GenAI can work in isolated pilots, but they’re struggling to translate that into repeatable, enterprise-grade performance. Agentic systems are the next step because they can reason, execute, and coordinate actions—not just generate content.
Q: Your announcement talks about “decoupling revenue growth from operating costs.” What does that mean in practice?
A: In the traditional enterprise model, scaling revenue often means scaling costs—more volume requires more people, more manual steps, and more operational overhead. Our objective is to decorrelate that relationship. With agentic systems embedded into high-volume processes, organizations can absorb growth in complexity and demand while keeping the cost base more stable. It’s not about replacing people—it’s about shifting human effort away from repetitive execution and toward oversight, judgment, and high-value work.
Q: Many companies are investing in AI already. Why aren’t pilots turning into production success?
A: “Pilot purgatory” usually happens for a few predictable reasons: unclear ownership, fragmented data, weak governance, and workflows that were never redesigned for autonomy. A chatbot demo is easy; an AI system that can take action safely inside finance, supply chain, customer operations, or industrial workflows requires permissions, auditability, exception handling, reliability metrics, and change management. Without that foundation, pilots remain impressive but non-scalable.
Q: You’ve said agentic AI is more than a tech upgrade. Why?
A: Because it’s a new economic lever, not just a new interface. Traditional automation has improved efficiency over time but has been challenged when dealing with complexity, whether it’s a new data model, unstructured data or systems upgrades. Agentic AI changes the unit of execution: systems can manage tasks end-to-end, coordinate across tools, and operate continuously within defined boundaries. That creates operating leverage—if the organization designs for it. If you treat agentic AI as a bolt-on, you’ll get isolated gains. If you redesign core processes around it, you can change the economics of growth.
Q: What’s different about Artefact’s approach?
A: Our programs are built around a simple idea: simplify before you agentify. Enterprises are complex—different systems, inconsistent processes, exceptions everywhere. If you agentify chaos, you scale chaos. We focus on structural integrity first, then autonomy.
Take the example of that large Retail client we are working with: their retail expansion process was a long, tedious one where Expansion team would feed their market research counterparts with hundreds of new projects and locations to be screened annually, creating a 3 months backlog. We could have come in and equipped the market research team with a GenAI tool that helped them scrape or collect competitive or economic data, a task that’s manual and cumbersome today. Instead, we revamped the entire process: the Expansion team can load any opportunity into a fully capable agentic solution that is going to perform an initial screening, discard about half of them and then send a pre-draft to the market research team for those that need further attention. Backlog is down to under a week now.
Q: Artefact’s slogan is that AI is about people. Can you expand on that?
A: Agentic AI programs are put in place to support and automate repetitive tasks. People remain the most important part of any organization, they remain the decision makers, and the creative engines. They determine which direction they would like the organization to go, and Agentic AI programs can make that possible. It’s the people and their willingness to embrace these programs that make or break organizational success. There’s also a human factor to an AI transformation: every tech-enabled transformation requires changing your team skillsets, operating models and mindsets. And that’s often a field that under-invested in most transformations I’ve seen.
Q: Can you walk us through the methodology?
A: We organize the work around four pillars:
- Process redesign & simplification We start by reducing complexity—standardizing and clarifying workflows so there’s a clean runway for automation.
- Strategic agentification Once the process is optimized, we deploy autonomous agents that can reason, make decisions within boundaries, and execute workflows end-to-end—while escalating edge cases.
- Data & infrastructure foundations Agents are only as reliable as the data and systems they operate on. We bring strong DataOps and MLOps discipline so agents are fueled by trusted, governed, and where needed real-time data.
- Operational leverage The goal is an operating model that redirects human potential from repetitive tasks to strategic oversight—supported by measurement, continuous improvement, and accountability.
Q: What kinds of enterprise processes are best suited to agentic systems?
A: The best candidates are high-volume, high-friction workflows where decisions are frequent but governable and where knowledge plays an important role - think customer operations, knowledge work, forecasting, supply chain, finance, marketing operations, and industrial processes. In practice, we look for areas with measurable KPIs, clear inputs and outputs, and a meaningful burden of repetitive coordination.
Q: How do you address risk, security, and compliance?
A: Governance isn’t something you bolt on at the end—it’s embedded from the start. That means controlled access to tools and data, audit logs of actions and decisions, clear boundaries for autonomy, and defined escalation paths when uncertainty is high. Production-grade agentic AI must be observable, explainable enough for the context, and accountable.
Q: You’re also introducing an “AI Quick Scan.” What is it?
A: The AI Quick Scan is an intensive diagnostic designed to help leaders determine where agentic AI will produce the highest ROI—and what it will take to deploy responsibly. In 4 to 8 weeks, we assess high-impact processes for agentic potential, quantify the expected “cost-to-revenue” decorrelation opportunity, and map the technical, data, and operating model requirements for rollout.
Q: What does success look like at the end of the Quick Scan?
A: Leaders walk away with a high-fidelity roadmap and a validated business case: prioritized use cases, expected value, risk considerations, required data and integration work, governance design, and an implementation plan that can move into delivery without guesswork. It also helps them align with their internal stakeholders (ExCom, Board, Shareholders) on the priorities and the investments required.
Q: If you could give enterprise leaders one piece of advice about agentic AI, what would it be?
A: Make big bets on core processes, not side experiments. Agentic AI creates real leverage when it’s applied to the workflows that run the business—supported by simplification, trusted data, and governance that enables scale.
About
Ghadi Hobeika is an experienced sales & marketing, data and e-commerce specialist with a strong background in digital media, growth hacking and analytics.
He is the Managing Partner & CEO, North America at Artefact, a global data-driven services company specialising in consulting for data transformation and digital & data marketing. Artefact is an independent global network with 36 offices in 27 countries, working with over 300 of major brands around the globe, such as Samsung, L'Oréal and Sanofi.
About Artefact
Artefact is a global leader in data and AI transformation, specializing in bridging data science and measurable business value. With a footprint across the US, Europe, and Asia, Artefact partners with leading brands to build the foundations of the AI-powered enterprise.
Copyright © 2026 California Business Journal. All Rights Reserved.
For California Business Journal Disclaimers, go to /terms-conditions/.
