TL;DR

Cloud data and AI consultancy Datatonic says organisations are undermining productivity by deploying AI in isolation from human workflows. The company argues that human-in-the-loop (HiTL) systems — where AI handles speed and scale while humans provide judgement and accountability — are where real enterprise value shows up.

The productivity gap

After years of AI investment, pressure is building on businesses to show returns. But Datatonic’s research indicates that many AI initiatives remain stuck in pilot stages due to limited user trust. Without proper integration into existing workflows, AI-powered insights fail to reach decision-makers, and efficiency gains never materialise.

“The biggest risk we see in the market is productivity leakage when AI exists in isolation from the people who actually run the business,” said CEO Scott Eivers.

Where HiTL is working

Datatonic points to agent-assisted software development as a clear example. Human teams decide what needs to be built, inspect requirements, and review plans. Once direction is set, AI agents construct modular components. The same pattern is appearing in finance and operations — AI-powered document processing has delivered a 70% reduction in invoice-processing costs in some cases, but finance teams still approve final outcomes.

“Humans create evaluation systems, validate plans, set guardrails, and make decisions. AI executes at speed and scale,” said CTO Andrew Harding.

The governance question

Many enterprises are failing to deploy fully autonomous agents safely, according to Datatonic. Security controls and governance frameworks are not keeping pace. Autonomy can only scale when organisations introduce approval checkpoints, benchmark performance standards, and implement evaluation systems as models evolve.

“Skipping governance doesn’t build speed, it creates risk,” Harding added.

Looking forward

Datatonic predicts significant acceleration in AI-assisted workloads over the next two years, with preparation and validation handled by AI agents. The company envisions expert departments run by smaller teams — finance, HR, marketing — each amplified by AI. The differentiator will be organisations that teach people to work with AI, not around it.