Lumel wants your budget to live where your data does

At Merck, the forecasting problem began with an arrangement familiar to anyone who has ever chased a budget: multiple people, multiple spreadsheets, and a central team trying to turn the contributions into a usable forecast. The pharmaceutical and life-sciences company wanted to automate that process. Its published Inforiver case study describes the replacement: legal P&L forecasting and variance reporting in Power BI, with inputs written back to Amazon Redshift. More than 100 users had been brought onto the solution when the case study was prepared.

There is a small but consequential idea tucked inside that description. The report becomes a place where people contribute information, rather than simply inspect it. A forecast can acquire a new number without first becoming an attachment. Lumel, the company behind Inforiver, has spent years building a business around that change in direction.

01 / The spreadsheet’s real rival is a round trip

Business intelligence has an asymmetry. It is good at presenting transactions that have already happened. Planning asks people to enter things that have not: next quarter’s hiring, a new product’s demand, a manager’s revised sales estimate. Those judgments are essential, yet they often live outside the system that measures the business.

Lumel makes enterprise performance management software, a category concerned with budgets, forecasts, reporting and the processes surrounding them. Its particular argument is architectural: build the planning layer on the customer’s data platform. Reuse the semantic model, the agreed definitions and relationships that make a Power BI report meaningful. Avoid rebuilding that knowledge in a separate planning database wherever possible.

Consider a hypothetical revenue plan. A sales manager changes an assumption, finance reviews it, and the revised forecast is saved alongside the data needed to compare it with actual sales. The practical gain is fewer handoffs. Whether the forecast itself improves still depends on the judgment of the people entering it.

02 / He sold the services and kept the software

Gopal Krishnamurthy started Visual BI in 2010. Consulting put the team close to the executives and operational managers using analytics, and the business developed products alongside its engagements. Lumel’s own account of that period describes a sequence of gaps: visualization, simulation, the governance of BI assets, and the distance between reporting and planning.

In 2021, Atos bought the Visual BI consulting business. Krishnamurthy and his partner, Jayaraman Anantharaman, retained the products. The independent Lumel company dates from that separation, although its experience reaches back to Visual BI’s beginning. A company can be young on paper and rather older in its habits.

Krishnamurthy subsequently described funding Lumel with proceeds from the services sale. A May 2025 Practical Founders interview reported an annual recurring revenue run rate above $12 million and a team exceeding 300 people. Those are historical interview figures, rather than an audited account of today’s business. They nevertheless explain the financing choice: the founder had capital to commit to a lengthy product effort without announcing a venture round.

The transition also changed the work. A consulting team can solve a customer’s exceptional problem through a custom engagement. A product company has to decide which exceptions belong in the software and which should remain exceptions. Krishnamurthy’s interviews describe that shift from services to a more self-service enterprise product business.

03 / A family of tools, with different jobs

Inforiver is the most direct way into the portfolio for a Power BI user. Reporting Matrix handles formatted management and financial reports, including paginated output and Excel or PDF exports. Writeback Matrix adds data entry, forecasting, collaboration and database writeback. Analytics+ combines charts, cards and tables for visual analysis and storytelling.

These are distinct products, and their commercial terms differ. Inforiver documentation describes SaaS and on-premise editions of Writeback Matrix, plus procurement through Microsoft AppSource or the Inforiver website. Someone evaluating a reporting visual should not assume they are buying the same deployment or subscription as a full planning application.

Elsewhere in the family, ValQ provides modeling and simulation in Power BI. xViz supplies specialized enterprise visuals, including Gantt charts and Performance Flow. The common thread is familiarity: the user remains near the reporting environment they already know, while acquiring capabilities it does not provide on its own.

04 / Three customers, three versions of the same nuisance

Adapa, a European packaging business, needed to streamline spreadsheet-based planning across 21 specialized operating centers. Its Inforiver case study describes distributed inputs, what-if simulations, rolling forecasts and centralized writeback, with commenting and an audit trail. The issue was organizational as much as numerical: the plan had many authors.

At ibex, the customer-engagement technology provider, forecasting needed frequent database updates. The published case study describes SQL Server writeback, support for 500 concurrent users, and roughly 38,000 writebacks accumulated by the time of publication. Its stated peak was one writeback a minute. That is a dated deployment result, not a universal performance guarantee.

Merck’s forecasting example adds the finance department’s view. Taken together, these cases show the market Lumel serves: business teams that need to collect judgments from many people while preserving a common reporting environment. The company reports more than 3,000 customer companies across its portfolio. That figure should not be mistaken for 3,000 installations of the newer Fabric capability.

05 / Microsoft moves the bet inside Fabric

The Microsoft relationship changed the distribution story. Lumel became a Fabric Strategic ISV Partner in 2024 and announced its native EPM workload in September 2025. In March 2026, the companies announced a co-engineered, first-party planning capability in Fabric IQ. Microsoft announced general availability in July, with billing starting August 1.

The integrated experience combines Planning, PowerTable and Intelligence. PowerTable deserves attention because plans frequently depend on entities that do not exist yet: a future product, a prospective customer, a project still awaiting approval. It gives business users governed applications for that reference data, with structured inputs, permissions and approval workflows.

Microsoft’s July announcement also described APIs and integration with deployment pipelines and Git workflows. Those details matter to the teams expected to operate the system after the demonstration ends. A finance-friendly interface still needs a manageable production environment.

“Our shared vision with Microsoft is to democratize planning for every organization that has invested in Power BI semantic models.”

Gopal Krishnamurthy · March 2026

06 / A session is longer than a click

The pricing pitch rewards a close reading. Fabric Planning consumes Fabric capacity through active user sessions. A session runs for 30 days; it is not a stopwatch recording a few minutes of editing. Roles matter, automation jobs meter separately, and SQL, OneLake and other Fabric workloads add their own consumption.

Lumel’s estimator gives illustrative pay-as-you-go session equivalents of $150 for a Planner, $30 for a Stakeholder and $6 for a Viewer, with lower reserved-capacity equivalents. These are estimates tied to capacity economics, not an all-inclusive software invoice. A seasonal reviewer can avoid year-round commitment, but an activated session still has a full window.

In August 2026, Microsoft refined the roles for PowerTable and Intelligence: editors could bill as Stakeholders when the item had no Planning sheets. That change makes the data-management use case less expensive than treating its editors as planning authors. It is a concrete example of the product’s economics changing after release.

07 / Copy the sequence, then test the assumptions

Lumel competes with established planning systems such as Anaplan, Workday Adaptive Planning, Oracle Cloud EPM and SAP Analytics Cloud, as well as Power BI-connected alternatives such as Acterys. Its clearest fit is an organization already invested in Microsoft analytics and looking to make planning part of that environment. The argument becomes less compelling when adopting Fabric would itself require a substantial platform change.

There is a useful evaluation sequence to borrow. Choose one recurring forecast. Agree on its measures. Identify who enters data, who approves it and where the results must be saved. Then test a complete cycle, including permissions, exceptions and capacity consumption. That is more revealing than counting features.

Reusing a semantic model only helps if people trust it. No-code tools cannot settle competing definitions of margin or decide who owns an assumption. Complex finance requirements also deserve explicit testing. The shorter data journey is attractive; the organization still has to decide where it is going.