Friedman Real Estate and Leni have announced that Friedman is building automated reporting agents and workflows on Leni’s infrastructure for commercial real estate. Friedman’s teams will use Leni’s platform to assemble the workflows needed, using their own data, their own definitions, and their own review steps. Those workflows now run on their own and produce reports Friedman can trace back to the source.
The two firms hold a shared view that a real estate company should be able to build its own automation instead of waiting on a vendor roadmap or a custom development cycle. As Friedman scaled its portfolio and took on new properties, its operations teams absorbed a growing volume of custom reporting requests, each one slightly different from the last.
Leni’s infrastructure gave Friedman a way to turn each of those requests into a repeatable agent, so the work runs without manual assembly, and the accuracy and audibility Friedman’s stakeholders expect stays intact.
Friedman’s first set of workflows covers custom reporting and the context around it. The data is pulled from Friedman’s ERP and accounting systems. It applies Friedman’s own context and definitions, produces reports and dashboards in Friedman’s format, exports them to PDF, and routes them for review. Work that used to be assembled by hand across several people and systems now runs as a workflow that Friedman controls and can change whenever a lender, owner, or internal team asks for something new.
“We did not want another reporting product that we would immediately need to change. We wanted to build our own workflows, and Leni’s infrastructure is what made that possible.
“Our team defined the metrics and the review steps, and the agents now produce the reports and dashboards on their own, in our format, traceable back to the source.
“When a lender or an owner asks for something new, we build it ourselves instead of filing a request and waiting.”
Jared Friedman, Co-CEO, Friedman Real Estate
As Leni supplies the layer underneath, Friedman did not have to fund a bespoke engineering build to get there. The connectors, data model, specialised models, and the checks were already in place. Friedman’s contribution was the part only the company knows – which numbers matter, how they are defined, what a finished report looks like and who signs off.
Leni turns what is normally a long systems project into three steps a team can work through on its own. Leni’s patent-pending Universal Data Model already speaks to seven major ERP and accounting systems, roughly 85% of the market, so teams connect their systems instead of specifying pipelines, mapping fields, or standing up a warehouse.
Leni’s patent-pending organisational memory captures the part no system of record holds – how the firm defines its metrics, what a finished report looks like, and who approves it. It builds as people work, so a team is not documenting its own processes up front.
With the data connected and the context in place, teams assemble their own agents, customisations, and workflows without designing the architecture underneath. Accuracy is handled by the platform – specialised models and verification loops check every step against the source before an output reaches a person.
“We provide the right infrastructure so that firms can experience the true promise of AI. Friedman’s data was locked in closed systems, and the knowledge of how Friedman reads that data lived with its people.
“We connect the systems and capture that context, then give teams specialized models and verification they can assemble into their own agents.
“Friedman built reporting agents first because reporting hurt the most. The point is that the context layer belongs to Friedman, and every workflow it builds next runs faster and cheaper on it.”
Arunabh Dastidar, Co-founder and Chief Executive Officer, Leni
Those checks are why teams can build without trading away reliability. In Leni’s published research, the architecture improved the accuracy and reliability of the underlying models by seven to 15 percentage points and cut the cost of reaching the result by roughly 65%. In production, it delivers task accuracy above 99.6% at about one-third the cost of frontier models, and every output carries a record of what was checked, against what, and when.
A packaged reporting tool is fixed on the day it ships, definitions change, owners ask new questions, portfolios take on new assets, and each change means another request routed back to a vendor. Infrastructure removes that dependency.
That context stays with Friedman, every connected system, corrected number, and approved deliverable becomes part of the firm’s own operating knowledge. Reporting was the starting point, and the same foundation supports budgeting and variance analysis, investor and board packages, document review, and broader operational work as Friedman extends it.
For the wider industry, the takeaway is that firms no longer need to construct this infrastructure in-house to build with AI. Leni supplies the data architecture, the context layer, the specialised models, and the guardrails. The firm supplies its own conventions and builds on top.
