The promise of artificial intelligence has become something of a corporate mantra. Every conference keynote echoes its potential and every earnings call touts its integration. Yet, a palpable frustration simmers beneath the surface. Companies are investing heavily in digital and AI transformations but the grand business outcomes—the leaps in efficiency, the surges in innovation, the definitive competitive edge—often remain elusive. The technology is there so why aren’t the results?
At the recent Gartner Application Innovation & Business Solutions Summit in Tokyo, Takashi Asaka, Head of DX Division at Ricoh Company, Ltd., pinpointed the core issue with refreshing clarity. The problem, he argues, is rarely the AI itself. It’s the business processes we’re asking AI to supercharge. We’re trying to bolt jet engines onto horse-drawn carts, expecting a smooth flight and then wondering why the ride is so turbulent.
Speaking from the trenches of Ricoh’s own multi-year transformation from a hardware-centric company to a digital services provider, Asaka outlined a radically different approach: AI Transformation (AX) through a process-first lens. This isn’t about chasing the shiniest new AI model; it’s about engineering the racetrack before you unleash the race car. His session, “Delivering Business Outcomes in the AI Era,” was less a sales pitch and more a masterclass in operational honesty, detailing how AI-ready processes, thoughtful human-AI collaboration, and purpose-driven deployment are the real keys to unlocking what Ricoh terms “Return on AI,” or RoAI.
Asaka began by grounding his talk in a stark reality, citing Gartner research indicating that Japanese companies, in particular, are struggling to meet CEO expectations for digital initiatives. Even with generative AI hailed as a universal productivity booster, only a limited number of firms see results that exceed expectations. The gap between investment and outcome is wide. Ricoh’s diagnosis? A missing foundation. For years, the company has internally championed Business Process Management (BPM), evolving it into a formalized “process DX” discipline. The goal was human-centric: to systematically eliminate the workplace stress born from tedious repetitive and error-prone tasks. This frontline-led cultural shift, supported by Centers of Excellence and armies of citizen developers, was the necessary groundwork. But as Asaka noted, this field-driven momentum also contains a hidden risk in the AI age.
When teams adopt AI tools in isolation—a department here automating reports a team there using an AI chatbot for customer queries—they often create new unforeseen inefficiencies. These siloed “islands of automation” can disrupt interconnected processes upstream and downstream. They also make AI-related costs spiral hidden across different budgets and tools. The solution, Asaka stressed, is to shift from ad-hoc AI adoption to a strategy that marries end-to-end process optimization with deliberate AI placement. This is the heart of maximizing RoAI. He broke this down into three interdependent pillars:
- Building AI-ready processes
- Intentionally designing human-AI collaboration
- Deploying AI with a clear bounded purpose
- Creating a clean core for smooth integration
- Reducing inefficiencies through orchestration
- Ensuring human accountability in AI operations
An AI-ready process is designed from the start with artificial intelligence as a core component not an afterthought. Asaka shared the example of Ricoh’s internal procurement operations. The company didn’t simply sprinkle AI onto its existing procure-to-pay (P2P) workflow. Instead, it first redesigned and standardized the entire process expanding automation to create a fully orchestrated pipeline. Only on this clean streamlined foundation was AI integrated into the System of Record environments that handle high-volume transactions. The result was a dual win: massively improved data processing efficiency and a significant reduction in the grueling error-sensitive manual work that plagued employees. This foundational work, often supported by platforms like Axon Ivy for process orchestration, creates what Asaka called a “clean core.” It allows AI to be integrated smoothly without the need for extensive and costly customization of legacy enterprise systems—a common pitfall that sinks many AI projects.
But technology alone isn’t the answer. Asaka was candid about the legitimate concerns surrounding AI reliability governance and accountability. Business operations especially in regulated environments demand accuracy and clear audit trails. Ricoh’s answer is a firm “human-in-the-loop” philosophy. In their model, AI handles the heavy lifting of data processing and pattern recognition but humans remain the final arbiters making judgment calls and retaining ultimate accountability. He illustrated this with an unexpected yet perfect example: streamlining invoice processing for the Ricoh Black Rams Tokyo rugby team.
| Aspect | Description |
|---|---|
| Invoice Management | Centralized application for processing invoices |
| AI-OCR Technology | AI enabling Optical Character Recognition for invoice items |
| Accounting Categorization | AI suggests correct categories for expenses |
| Role Shift | From manual data entry to final review and approval |
| Outcome | Reduced workload fewer errors and shorter processing time |
This leads to the final pillar: purpose-driven AI deployment. Asaka emphasized that not all AI is created equal. A large language model might be terrible at precise data extraction but excellent at summarizing documents. The key is selecting the right AI for the right task within a well-designed process. This precision improves transparency ensures explainability and crucially helps control costs by preventing the misuse of expensive generalized AI on tasks better suited to simpler cheaper tools.
The most compelling part of Ricoh’s story may be how it turns this internal discipline into customer value. The company now uses AI to design and improve business processes themselves employing tools like process modeling AI that can generate workflow diagrams from conversations. This accelerates the journey from problem identification to solution implementation. Asaka highlighted that Ricoh’s continued investment in human expertise—its business analysts and process consultants—is amplified not replaced by AI. Their deep knowledge combined with these new tools dramatically shortens the time to deliver tangible results for clients.
Asaka concluded with a reflection on the nature of business particularly in Japan. The processes are often famously detailed and complex. While this can be seen as a hurdle for transformation Asaka reframed it as a potential strength. That precision when optimized and paired effectively with AI becomes a formidable source of competitive advantage. By starting with the process organizations move beyond simply doing AI to achieving AX—a fundamental transformation in how work is conceived and executed. Ricoh acting as its own “Client Zero” demonstrates that the path to fulfillment through work in the AI era isn’t found in a flashy algorithm alone but in the deliberate human-centric engineering of the work itself.
