RUNBIRD works where operations, technology, data, and business decisions stop lining up. We investigate the real problem, establish what the evidence supports, and design or build systems that make work and decisions better.
Our starting point isn't a preferred software stack or industry. It's a business problem with consequences. We move between investigation, engineering, and assurance depending on what the evidence requires.
When reports disagree, models have changed hands, or operational results can't be explained, we trace the logic back to the people, processes, sources, and decisions.
When the investigation reveals a broken workflow or missing capability, we design and implement practical systems—not technology for its own sake.
Before a critical report or system change is relied upon, we document definitions, compare versions, test important outputs, and expose gaps requiring approval.
ModelIntel is RUNBIRD's evolving investigation workbench for inherited Power BI and Excel environments. It helps organize model structure, definitions, relationships, changes, tests, and evidence in one place.
The broader direction is an ontology-driven workspace connecting systems, claims, dependencies, people, decisions, and proof—not another static dashboard gallery.
Current state: developed local pilots and demonstrations. Live cross-system autonomy, comprehensive numerical certification, and enterprise production readiness are not claimed.
The common thread in my work is reconstructing how a business actually functions, fixing what gets lost between systems and teams, and making the next decision easier.
Six actual experience stories, explorable problem → investigation → intervention → outcome. No fake client wins or proprietary screenshots.
Analyzed leasing, tenant mix and location performance across a U.S. retail portfolio of more than 90 mall properties. Work included combining operational and movement signals, including Wi-Fi and camera-derived activity data, into more useful analytical views.
Location intelligence · Multi-site performance · Operational data integration
Inspect the work →Built and maintained rent-roll quality checks, pricing and lease-velocity analysis, exposure tracking, Yardi-based reporting processes and executive visibility across a multi-property housing portfolio.
ERP / PMS normalization · Revenue assurance · Inherited Power BI systems
Inspect the work →Helped an independent Chicago salon put the operating pieces together: website and online store, press visibility, customer and stylist analysis, pricing research, STX point-of-sale reporting and in-store improvements, including an ATM and a custom photo setup.
Small-business operations · Customer economics · Commerce and workflow design
Inspect the work →Worked on contracting and decision workflows where turnaround time, document movement and inconsistent processes directly affected how quickly teams could complete work. The focus was finding bottlenecks and engineering more reliable handoffs.
Process automation · Approval workflows · Operational systems engineering
Inspect the work →Investigated and reconstructed operating financial and P&L reporting in a complex housing and construction environment, following differences in interest treatment, source logic and reporting periods back to their origins. The work centered on explaining movements and rebuilding a consistent reporting basis—not accepting the headline total at face value.
Financial reporting diagnostics · ERP reconciliation · CFO reporting confidence
Inspect the work →Worked across inherited Power BI, SQL Server, Dynamics and Yardi reporting to reconstruct calculations, investigate data relationships, and maintain historical context. This included using time-aware source records to explain why the same operational question could produce different answers across periods.
Power BI takeover · Model lineage · Temporal reporting · Governance handoffs
Inspect the work →Different industries. Same discipline. If your reports disagree, your systems don't connect, or no one can explain the inherited logic, start with a scoped investigation.
Discuss a similar problem ↗Not every problem needs a long consultancy project. Some need a diagnostic. Some already have a clear fix. Some need independent evidence before leadership can rely on the result.
A bounded investigation of inherited reporting, contested definitions, and change risk. One available first-pilot scope starts at $1,250, subject to fit and written scope.
Trace a cross-team or cross-platform business problem, identify failure points, and specify the sequence of fixes that would matter most.
Automate manual processes, connect workflows, build focused applications, or rework data and decision pipelines after requirements are established.
Check important numbers, dependencies, model changes and documentation before a release, handoff or recurring executive reporting cycle.
Two reconstructed investigations: financial reconciliation and inherited report assurance. Original fixture data, difference bridges, and explicit HOLD decisions.
Three interactive, fictional investigations show how a reporting conflict, a risky model change, or a broken operational handoff becomes an evidence trail and a defensible next decision. No invented client results.
Four fictional source files → recorded Python analysis → record-level differences, mapping conflicts, definition warnings and a release HOLD. Includes reproducible source data and hashes.
Two source exports have identical totals. But two individual records disagree by $20 in opposite directions. A grand-total check falsely suggests everything is fine.
Open the evidence example ↗Fabricated demonstration data only. None of these examples represent customer outcomes or live-environment certification.
What is failing, what consequences does it create, and what would count as resolution?
Trace data, logic, workflows and ownership. Separate observation, hypothesis, and demonstrated fact.
Report the findings, implement the fix, or establish repeatable assurance—with boundaries and acceptance criteria.
RUNBIRD is led by Manoah Williams, an analytics and operational intelligence leader whose work has crossed retail real estate, multifamily operations, contracting, independent business and enterprise reporting. His earlier service as an Army Military Intelligence Sergeant shaped a working discipline: connect the signals, distinguish evidence from assumptions, and make the next decision explicit.
The fix might be a semantic model, a reliable operational process, an automated workflow, a focused application or a better-designed physical handoff. Tools matter, but the outcome is the product.
Describe the problem and why it matters. We'll discuss whether the right next step is an investigation, an engineering project, or an independent validation.
Prefer email? Send directly to manoah@runbird.io ↗ No forms or attachments required.