Enterprises today are not short of data. What they lack is the ability to act on it decisively. Despite years of investment in business intelligence tools, dashboards, and analytics platforms, the average enterprise decision success rate remains stubbornly low. Bengaluru-based startup DecisionX AI believes it has identified the structural gap—and built a platform to close it.
The startup was founded by Ranjan Kumar, who earlier founded Entropik, Emotion AI company and upon exiting it in 2024, started DecisionX. It was launched in February this year and has accumulated five to six months of customer adoption, with what the company describes as encouraging early traction.
The product is pitched not as another analytics layer, but as a causal decision intelligence platform, one that converts enterprise data into actionable, auditable decisions at scale.
“Raw data is no longer a differentiator,” Kumar says. “Judgement and decision quality are the new competitive advantages for enterprises. The question isn’t whether you have the data—it’s whether you can act on it correctly and consistently.”
The Limits of Semantic AI
The enterprise AI landscape has evolved rapidly, from early attempts to connect large language models directly to databases, to the now-common approach of building semantic layers, data dictionaries, taxonomies, and metadata models, to give AI systems context. However, Kumar argues that semantics alone reach a ceiling when applied to real decision-making.
Semantic models are adept at answering the ‘what’ questions: what are my sales figures or what did revenue look like last quarter? But enterprise leaders rarely make decisions based solely on descriptive analytics. They need answers to the ‘why’ and ‘what if’ questions: why did sales decline in a particular region, and what happens if marketing spend is shifted to a different segment? These causal questions, Kumar contends, are precisely where most semantic-only approaches begin to fall apart.
“Naive LLM-to-data approaches succeed at small scale but largely fail in enterprise proofs of concept,” he explains. “The missing ingredient is a causal understanding of the business, not just what the data says, but what is actually driving outcomes.”
The Causal Ontology Differentiator
To address this gap, DecisionX’s developed Ontology Studio, a proprietary layer that sits atop a conventional semantic model and continuously learns cause-and-effect relationships from enterprise data. Rather than requiring data teams to manually codify relationships, the platform builds and updates a causal fabric autonomously as new data arrives, enabling forecasting, root-cause analysis, scenario planning, and optimisation that static semantic models cannot support.
“We built a causal ontology that continuously learns cause–effect relationships on top of semantic context,” Kumar says. “This is what lets us answer not just what happened, but why it happened and what you should do about it.”
This self-learning architecture is also designed to address one of the most persistent failure modes in enterprise AI deployment: the context layer becoming stale. When business structures, hierarchies, or data definitions change, many platforms require manual intervention to stay accurate. DecisionX says its model updates automatically, reducing the maintenance burden that has historically derailed production-grade deployments.
Enterprise-Grade Governance by Design
Beyond its causal reasoning layer, DecisionX has built a suite of technical and governance features to address enterprise adoption challenges that cause AI pilots to stall before reaching production. These include mandatory data-led approvals before any reasoning chain is acted upon, full traceability to back to original data sources and the specific code invoked during analysis, auditable reasoning chains that allow compliance and risk teams to understand how a conclusion was reached, and enforced human-in-the-loop controls designed for regulated industries such as pharmaceuticals and financial services.
“Governance, auditability, and human-in-the-loop controls are central features, not afterthoughts,” Kumar notes. “In regulated environments, you cannot deploy AI that cannot explain itself. Every decision the platform surfaces can be traced back to its source.”
The underlying production-scale architecture underpinning the platform is designed to handle enterprise data volumes without degradation, another area where Kumar believes many platforms succeed in controlled pilots but struggle under real workloads.
From Months to Seconds
The platform's practical impact is perhaps best illustrated through deployment timelines and customer outcomes. Using prebuilt data agents, DecisionX brings new customers live in approximately 15 days after receiving access to their data—a sharp contrast to the months-long implementation cycles that have traditionally characterised enterprise analytics projects.
Once deployed, the compression of analytical workflows is significant. Tasks that previously required an analyst to stitch together multiple dashboards, build presentation decks, and synthesise findings over days or weeks are transformed into conversational reasoning sessions capable of producing root-cause analyses, scenario outputs, and presentation ready reports in a fraction of the time.
“The platform reduces a typical 30-day analyst forecasting project to interactive outputs in roughly 45 seconds,” Kumar says. “That’s not an incremental improvement, it’s a fundamentally different way of working.”
A Pharma Case Study
One of DecisionX’s early customers is a large Indian pharmaceutical company, where the platform has been applied to brand performance analysis across multiple geographies and stock-keeping units. Previously, each brand required approximately 30 days of analyst effort to conduct a root-cause investigation into performance changes. The platform now performs these analyses continuously while flagging causal factors that human analysts had overlooked.
In one instance, the platform identified prescription behaviour as the primary driver of a performance issue, rather than the marketing spend metrics that the team had been focused on. This shift in causal attribution changed the remediation strategy and delivered measurable impact within a quarter of deployment.
The platform also records decisions alongside their causal context, enabling teams to search historical decisions, track post-decision impact, and build an institutional learning loop, incrementally raising the organisation’s decision success rate over time.
A Decision Signature for Leaders
Looking further ahead, Kumar has outlined plans for a “decision-signature” product, a feature designed to model individual decision-making styles by analysing the questions users ask and the reasoning patterns they apply. The platform would capture whether a user tends to reason tactically or strategically, and allow other employees to query the system in a mode that emulates a particular leader’s decision style.
“The primary enterprise use case is scaling a leader’s decision competency so other employees can think like that leader,” Kumar explains. “Every person has a decision signature—a pattern of how they approach problems. We want to make that transferable.”
The product remains early-stage, with no announced launch date, but is being developed in parallel with ongoing customer engagements from which the company plans to extract case studies.
The India Advantage
DecisionX is deliberately building from India, leveraging what Kumar describes as strong engineering talent, cost advantages, and fast access to enterprise proof points through the domestic market. The strategy is to build a credible portfolio of outcomes with Indian enterprises before scaling globally, a sequencing choice that also reflects the current limitations of India’s go-to-market infrastructure for enterprise software on the world stage.
Kumar positions the company within a broader evolution of the analytics market: from static dashboards to conversational business intelligence and now to causal decision infrastructure, platforms that do not merely surface information but materially improve the rate at which enterprises make correct, impactful decisions.
“Decision intelligence has evolved from dashboards to causal infrastructure that can accelerate enterprise decision velocity and quality,” he says. “We’re not selling a tool. We’re owning the outcome.”
Your reaction