At a global consumer health company, demand planning, procurement and logistics had a problem that would be familiar to anyone who has survived a meeting about a spreadsheet. The same metric meant different things in different systems. Before anyone could decide what to replenish, someone had to assemble the data and reconcile the answers. The machinery worked. The meanings disagreed.
- Intuitive.ai builds enterprise AI, data and cloud systems.
- Its IntuitiveIQ product adds shared business definitions and controls to existing technology.
- The revealing test: whether people can use the answers to make decisions.
Intuitive.ai describes a six-week project that tackled this disagreement inside the customer’s existing Databricks environment. Planners ultimately rated 87% of insights ready for decisions, above an 80% target. That is a useful way to judge an AI project: put the answer in front of the person who must live with it.
The company beneath the answer
Intuitive.ai sells the engineering required to make enterprise technology useful. Its services span data and AI, application and database modernization, cloud platforms, cybersecurity and networking. Customers bring complicated estates: old applications, scattered information, security requirements and operations that cannot simply pause while somebody installs the future.
Its public customer testimonials include SS&C Intralinks, Aptos, AmeriHealth, Hutchinson, Baptist Health and Santander US. These are organizations buying implementation, integration and operational responsibility. For a bank, that might mean modernizing data infrastructure. For an industrial business, it might mean connecting operational signals to systems that can support analysis and automation.
“Your willingness to say ‘I don’t know’ and then own the work of finding answers.”Steve Roy, SS&C Intralinks, in a company-published testimonial
The distinction matters in a market crowded with convincing demonstrations. A demonstration can begin with tidy inputs. A customer begins with whatever survived the last acquisition, migration and reporting deadline. Intuitive’s opportunity sits in that awkward distance between a promising model and a system people trust enough to use.
A cloud company changes its question
The business began as Intuitive Technology Partners in 2012. Its published history records an eight-country data-center migration for an insurer in 2013 and a first major public-cloud migration on AWS in 2016. Azure and Google Cloud followed in 2019. The Intuitive.Cloud name arrived in 2020.
A Data and AI practice launched in 2021. The company adopted Intuitive.ai in 2025 and lists the launch of IntuitiveIQ in 2026. Jay Modh is founder and CEO; Indraneel Shah is co-founder and COO. Read together, the milestones describe a supplier moving progressively closer to the decisions its infrastructure supports.
The progression offers a plausible explanation for its positioning. Moving a workload teaches you about dependencies. Running it teaches you about failure. Adding AI introduces another dependency: whether the organization agrees on what its information means. That connection is an interpretation of the company’s trajectory, rather than a claim that its early migrations foretold its current product.
Give the machines a dictionary
IntuitiveIQ is the product expression of this argument. It supplies a shared knowledge layer and an agent control plane: business meaning underneath the answer, with rules governing what software agents may do next. The product describes evidence trails, explainability, access controls and human approvals where required.
Its promise is to work across existing models, tools and data. An enterprise can ask questions in ordinary language while keeping the definitions, permissions and reasoning attached. Think of the knowledge layer as an agreed dictionary that software can consult. The control plane then determines which actions require review and which can proceed.
This makes Intuitive a combination of engineering services and proprietary software. Its competitive argument rests on connecting those pieces across the customer’s estate. Buyers can also assemble internal teams, hire specialist consultancies or use larger systems integrators. The sensible comparison is the work each option will own after the presentation ends.
The model was missing the customers
One telecom case makes the operational problem unusually clear. Existing churn models missed too many subscribers who were likely to leave. Small input changes disrupted performance, and business users struggled to understand the predictions. Slow, disconnected pipelines made the whole arrangement harder to maintain.
Intuitive says it rebuilt the system with the client’s data science, engineering and marketing teams. The work included redefining baselines, simplifying features, removing information leakage and adding explanations through LIME, SHAP and SageMaker Clarify. Automated deployment, retraining and drift monitoring completed the surrounding machinery.
The company reports nearly twice the benchmark model performance, especially on metrics for imbalanced data. That result should be read as a case-specific vendor report. The transferable lesson lies in the repair: inspect the data and operating process alongside the model. A prediction needs an explanation, a delivery mechanism and someone able to act.
Speed has an operating bill
In a separate insurer project, Intuitive built a questionnaire-led climate vulnerability application using internal datasets, vulnerability maps and Claude. It reports recommendations within five minutes and a 90% reduction in assessment costs through reduced manual work. Those figures describe the assessment workflow; they do not establish the price of buying the system.
Operations remain part of the offer. On September 14, 2026, Intuitive announced AWS Managed Service Provider Specialization following technical validation and an independent audit. Its managed services cover cloud operations, security, FinOps and monitoring. On October 7, it announced Preferred status in Anthropic’s Claude Partner Network.
Start with a question worth settling
The supply-chain project suggests a practical buying discipline: agree on the questions and the scorecard before asking software to answer. Bring the actual decision makers into validation. Keep the first scope small enough to inspect, then earn a broader deployment through results.
That approach depends on accessible data, owners who can settle definitions and users willing to judge the output. Without those conditions, another interface merely adds another place for disagreement. The interesting thing about Intuitive.ai is how much of its AI proposition depends on an old management skill: getting people to agree what they mean.