Axium Industries Navigates AI Agent Challenges

Singapore-based industrial AI firm Axium Industries has successfully navigated a market where many AI agent projects have faltered. The company specializes in building AI agents for industrial clients, concentrating on data connectivity and analysis rather than conventional application development.

According to Jeremy Ang, co-founder and CEO of Axium, the company dedicates 80% of its time to data work and only 20% to AI. This approach has enabled clients to identify capital leakage within their procurement, inventory, and resource allocation processes. Axium Industries has experienced significant growth, increasing its projects from two to 19 and seeing annual contract sizes grow from six to seven figures in USD.

Co-founders Tobias Leong and Jeremy Ang draw inspiration from Palantir's Forward Deployed Engineering (FDE) model. Their team dispatches engineers to client factories to address real-world problems. Ang and Leong emphasize that the bulk of the effort lies in data engineering, not solely in developing AI algorithms.

Axium's success contrasts with broader trends, as an MIT Media Lab report indicates 95% of AI pilot projects have failed to deliver measurable returns. Gartner predicts that over 40% of agentic AI projects will be canceled by the end of 2027, citing escalating costs, unclear business value, and inadequate risk controls.

The company quickly achieved product-market fit, becoming profitable within its first six months. Axium's client base has expanded from the oil and gas sector to include manufacturing, maritime, retail, and aerospace industries, demonstrating the applicability of its strategy across diverse sectors.