Forbes contributor Dustin Johnson (CTO at Seeq) outlines Decision Intelligence (DI) best practices for industrial organisations, diagnosing siloed sensor streams, maintenance logs and operator notes as barriers to timely decisions, and noting the loss of institutional knowledge as experts retire. The Forbes article cites a survey in which more than a quarter of data and analytics teams estimate annual losses above $5 million, and 7% estimate losses of $25 million or more. The piece presents DI as a discipline that digitizes and models decisions as assets to close the insight-to-action gap, and lists core capabilities such as contextualizing data and capturing knowledge for reuse. The article positions DI as an integrated decision layer for operations rather than mere automation, stressing human-AI-knowledge convergence as the objective.
What happened
Forbes contributor Dustin Johnson, identified as CTO at Seeq, published a practitioner-oriented guide on Decision Intelligence (DI) best practices for industrial organisations on May 28, 2026. The article reports that industrial teams are contending with disconnected sensor streams, maintenance logs, lab results and operator notes, alongside the retirement-driven loss of tacit expertise. Forbes cites a survey finding that more than a quarter of data and analytics teams estimate annual losses above $5 million, and 7% estimate losses of $25 million or more. The piece presents a working definition of Decision Intelligence as "a practical discipline that advances decision making by explicitly understanding and engineering how decisions are made and how outcomes are evaluated, managed and improved via feedback," and enumerates capabilities such as contextualization of data and knowledge capture.
Editorial analysis - technical context
Industry-pattern observations: Implementing DI in industrial settings commonly requires integrating multimodal telemetry with enterprise metadata and operational context. Typical technical building blocks include durable context layers, canonical ontologies, lineage and provenance tracking, and tooling to capture tacit knowledge (runbooks, operator notes, postmortems) as machine-readable assets. These components enable repeatable decision logic and feedback loops that connect outcomes back into predictive or prescriptive models. Practical challenges observed across deployments include schema and time-alignment of streaming signals, managing label and concept drift, and instrumenting the human-in-the-loop steps for continuous learning.
Context and significance
For practitioners, DI reframes work from isolated reporting to engineering decisions as first-class artifacts. That shift raises engineering priorities around data quality, metadata, observability and knowledge capture rather than only model accuracy. Organizations that centralise decision context and codify expert judgement can reduce ramp time for new operators and preserve institutional memory, which becomes especially relevant as experienced staff retire.
What to watch
- •Adoption signals: product features explicitly supporting decision models, decision catalogs, and knowledge-capture workflows in operational analytics vendors.
- •Technical indicators: investments in canonical ontologies, provenance, time-series alignment tooling, and feedback instrumentation for outcome evaluation.
- •Practitioner metrics: reductions in decision latency, repeatability measures, and the degree to which outcome feedback is automated into model or rules updates.
Bottom line
The Forbes guide maps DI concepts to concrete industrial pain points and capability areas. It emphasises integrating human expertise, data, and analytics into a continuous decision-improvement loop rather than treating DI as pure automation.
Scoring Rationale #
This is a practical, practitioner-focused best-practices piece that clarifies Decision Intelligence needs for industrial operations. It is useful for teams designing operational analytics and knowledge-capture workflows, but it is not a technical breakthrough or major product launch.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.