Tiankai Feng, Data & AI Strategy Consultant, author of Humanizing Data Strategy and Humanizing AI Strategy, and Vice President at DAMA Germany, spoke at Big Data & AI World Frankfurt 2026 about how organisations can scale AI and generate business value.  

Only 4% of companies are creating substantial value with AI, according to one of the figures Tiankai Feng shared at the start of his session. He also cited figures suggesting that 87% of AI projects never make it into production and 81% of workers still don’t use AI in their day-to-day work. 

Feng’s keynote looked at six areas that he identified as key to scaling AI in a valuable way: foundational architecture, operating model, ready data, experience for humans and AI, strategic alignment and trustworthy AI.

His section on strategic alignment focused on how organisations choose which AI projects deserve time, funding, and resources. 

Treat AI Investment as a Portfolio 

Feng started with the goals a business already has across strategic, tactical, and operational levels. AI initiatives can then be assessed against those objectives, with teams estimating their potential value and the effort involved.

“Based on how much effort it would take and how much value we believe it would give us, we can then build a portfolio”, he said. 

That portfolio needs regular attention. Feng described a lightweight planning and governance process in which organisations check whether projects are delivering what was expected, then decide whether to continue them, expand them, or stop them.  

The ability to stop matters because a place in the portfolio does not guarantee continued funding. 

Look Beyond One KPI 

Feng’s approach to measurement links what happens technically with what happens in the business. 

He described technical and operational measures at one end, and business and strategic measures at the other. His suggestion is to think in terms of KPI trees, in which those measures influence each other. That gives teams a clearer route into root-cause analysis when the final result is disappointing.  

It also makes success harder to reduce to a single headline number. The business result sits at the end of a chain of connected measures. 

When Experimentation Reaches Production 

Strategic alignment also brings three balancing acts into view: 

• Experimentation vs production  

• Refinement vs redesign  

• TCO vs value realisation  

They sit at a practical point in the journey from a promising use case to something the organisation is prepared to run. Feng’s framework asks teams to consider how an initiative moves into production, whether existing ways of working need refinement or redesign, and how total cost of ownership compares with the value being realised. 

Trust Has a Human Side 

Trust is another part of Feng’s framework, and he treats it as more than a governance checklist. 

“Trustworthiness itself is a very human emotion. So I can trust something while others might not trust it”, he said. 

Later in the session, he returned to the organisational side of trust. AI governance, he explained, can involve many people working together, which makes the distinction between accountability and responsibility important. Without clear ownership, problems can quickly lead to finger-pointing.  

Four Questions to Define Your AI Purpose 

Near the end of the session, Feng brought the discussion together with four questions: 

• What is exciting to do with AI?  

• What are we allowed to do with AI?  

• What can we technically do with AI?  

• What does our business strategy need from AI?  

He places AI purpose at the intersection of those four areas. Feng also recommends an early “handshake” between the people responsible for them, so an initiative reflects those needs from the beginning and is less likely to hit blockers later.  

Further reading: Humanizing AI Strategy by Tiankai Feng.