Tackle Enterprise AI’s Hardest Question At Forrester’s AI Forums

Every organization is investing in AI. Far fewer are getting business outcomes from it. Forrester’s Customer Experience Index (CX Index™) scores remain mediocre, AI-related costs are growing faster than the value they create, and the gap between experimentation and execution is widening. That’s the hurdle we’ll help you overcome at Forrester’s AI Forums this year, happening in Singapore on August 20 and Sydney on August 25.

For the first time in the region, we’re bringing together four tracks spanning the enterprise AI agenda: B2B leaders confronting the GTM singularity, CX leaders building the experiences AI can’t replicate, security leaders adopting AI with the controls to back it up, and technology leaders architecting the operating model that makes all of it work.

For attendees focusing on technology, what can you expect in August? My team and I will tackle what it takes to architect context and intent. In this tech-focused session track, CIOs, CTOs, enterprise architects, and heads of AI will consider the hardest question in enterprise AI right now: Why does task-level AI productivity fail to compound into firm-level business value?

The Cognitive Operating Model For The AI Era

Leslie Joseph will open the track by introducing Forrester’s cognitive operating model, a reference framework that offers an architectural blueprint for organizational reinvention in the age of AI. Enterprise operating models were built for the digital era: human-only workforces, application-centric architectures, and process logic encoded in deterministic workflows. Those assumptions collapse when AI agents need to act with purpose, side by side with humans, and within the guardrails, standards, and ethics established inside your organization.

Skills Are The New Building Blocks — For Humans And Agents Alike

Once you see the operating model as the constraint, the next question is: What’s the unit of work in this new model? We saw at last year’s event that focusing on use cases is a trap, failing organizations that can’t scale beyond disconnected pilots. The atomic unit of value is the skill: a business-defined cognitive unit of work that transcends considerations of who (human or AI) performs the work and how its fulfillment is architected.

Sam Higgins will make the case for a skills-oriented agentic architecture that enables enterprises to establish a shared language for human and AI capabilities, manage them as reusable products, and give organizations the composability they need to govern agentic work without strangling it. If service-oriented architecture (SOA) gave us the building blocks for digital, this new skills-oriented agentic architecture (SOAA) does the same for the cognitive era; the organizations that get skills right will scale. Organizations furthest along with use-case-driven AI may find the pivot hardest because their governance structures and team shapes were built for a model they now need to unwind.

Your Agents Are Only As Good As The Context You Give Them

Even with the right operating model and the right skills architecture, agentic AI stalls if agents can’t access trustworthy organizational context. And right now, most can’t. Context rot, noisy retrieval, and semantic ambiguity across policies and data are all scaling walls, producing the variance, hallucinations, and compliance risk that destroy executive confidence. A session from Charlie Dai will map the evolution from prompt engineering to context engineering and beyond: the semantic layers, knowledge graphs, and governance frameworks that create the reliable foundation that agentic systems need at enterprise scale. It’s less glamorous than the strategy conversations, but if I had to pick one session where the gap between “We’re doing AI” and “AI is working” gets closed, it’s this one.

Then, We’ll Put It All Into Practice

The track closes with a two-hour interactive workshop that Leslie Joseph and I will lead. We’ll move beyond frameworks and into workflow reinvention — working through hands-on exercises that apply four modes of AI-driven work (augmentation, automation, enrichment, and reinvention) to real processes. The goal: Leave with a concrete method for connecting AI initiatives to business outcomes, evaluating where agentic AI creates the most impact, and identifying the governance and change considerations that matter when you move from experimentation to operating model reinvention.

Four Tracks. One Forum. One Imperative.

The tech track is one part of a broader conversation, and connecting the four tracks is the same imperative: AI must move from a technology initiative to an enterprise commitment with executive ownership, cross-functional governance, and measurable business outcomes.

Attend Forrester’s AI Forum Singapore (August 20) or AI Forum Sydney (August 25) this year, because if you’re a technology leader past the pilot stage and ready to close the gap between AI activity and AI outcomes, this is where you want to be.