Selected work

Outcomes, not overhead.

A look at systems we designed, built, and still run in production, across short-term rentals, property management, real estate, finance, and operations. Client details are kept confidential, so these are described by the work and the result.

Engagements anonymized · outcomes are real and in production
40+
Services running
in production
20+
Platforms wired
into one system
7+
Self-hosted
MCP servers
100%
Code & infra
the client owns
Case studies

Problems we took off the table.

Each of these started as manual work, glue code, or a tool that did 80% of the job. We built the missing 20% and connected the rest.

Short-term rental

Books that close themselves

01 Challenge

A multi-property STR operator ran revenue across booking channels, a bank, and spreadsheets. Monthly owner statements took days to assemble by hand and rarely tied out to what actually hit the account.

What we built & the result

Automated owner statements, channel-payout reconciliation, and guest-comms, so statements generate in minutes, payouts match themselves, and routine guest messages send without a human.

GuestyBaselaneOwner statementsPayout reconciliationGuest comms

Property management

A growing portfolio, one source of truth

02 Challenge

A long-term rental management firm tracked rent roll, delinquencies, and owner reporting by hand across disconnected tools. Partners only saw the numbers when someone built a report.

What we built & the result

A live integration plus a self-contained monthly outstanding-balance report that emails the partners on schedule: balances visible at a glance, owners self-serve, month-end stops being a project.

DoorLoopRent rollDelinquency reportingOwner portaln8n

Real estate operations

Due diligence at the speed of a query

03 Challenge

Property research, ownership, comps, and public records were pulled by hand across a dozen sites for every deal: slow, inconsistent, and easy to leave half-done.

What we built & the result

Unified records tooling over property, owner, comps, and court data, plus an e-signature pipeline, so a full diligence dossier comes together in minutes and documents go out for signature the same day.

ATTOMCourt recordsSkip-tracee-SignatureDocument automation

Finance & bookkeeping

Money that reconciles itself

04 Challenge

Card spend, bank activity, and the general ledger drifted apart between platforms. Every month-end meant re-keying transactions and hunting for receipts.

What we built & the result

Bank-to-ledger sync with automated card-purchase import, receipt capture, and rules-based categorization, so transactions land categorized with receipts attached and reconciliation becomes review, not re-entry.

QuickBooksBaselaneReceipt captureAuto-categorizationReconciliation

AI automation

An ops team that never sleeps

05 Challenge

Repetitive inbox and text triage, status-chasing, and routine code fixes were quietly eating real staff hours every week.

What we built & the result

A self-hosted MCP gateway connecting AI to live business systems, an autonomous issue-fixing bot, and human-in-the-loop inbox/SMS triage, so routine messages draft and route themselves, code issues get fixed and opened as PRs, and people only touch the exceptions.

MCP infrastructureAutonomous agentsInbox triageSMS triageclaude-hub
Public example of the MCP work → vin-mcp on GitHub ↗

Records & compliance

Answers straight from the source

06 Challenge

Court, registry, and regulatory lookups were manual, slow, and easy to let go stale between checks.

What we built & the result

Live integrations to court systems and federal dockets, records decoders, and scheduled monitors, so authoritative answers are on demand and watched continuously instead of checked once in a while.

VA CourtsPACER / RECAPVIN decodeSearch ConsoleScheduled monitors
Who we work with

Sectors we know cold.

Most of our work clusters where operations get messy and off-the-shelf tools run out of road.

Short-term rental operators Long-term property management Real estate & deal operations Bookkeeping & finance Professional & legal services Government & commercial workflows Multi-entity operators Owner-operated small business
About this page

Questions about the work.

Why are the case studies anonymized?

Because that is the condition of doing the work. Naming a client, a portfolio, or a dollar figure would break the confidentiality NLMA promises every client, so each study leads with the problem and the outcome instead. A named version does exist, with real clients, real numbers, and the projects currently in flight; it sits behind a short non-disclosure agreement you can sign online and read in the same sitting. The public page is redacted, not padded.

What kinds of systems does NLMA build?

NLMA builds the connective tissue that sits between tools a business already pays for. On this page that means owner statements and channel-payout reconciliation for a short-term rental operator, live rent-roll and delinquency reporting for a long-term manager, diligence tooling that joins property, owner, comps, and court records, bank-to-ledger sync for bookkeeping, and self-hosted MCP infrastructure that lets AI agents act on live business systems. The 12 client sites and products listed further down are also NLMA builds, still running.

Can I see the actual code?

Yes. Most of what NLMA builds runs privately on client infrastructure, but the public sample is real and readable: 20 repositories under the NextLevelManagementAdvisors organization on GitHub, including vin-mcp, vacourts-mcp, turno-mcp, rag-mcp, github-mcp-auth, and gsuite-skills-plugin. They are the Model Context Protocol servers and supporting tooling that the automation in these case studies depends on. Every one is licensed AGPL-3.0, with commercial terms available on request. Source code is a harder test to fake than a case study, so judge the engineering before you sign anything.

Last reviewed: .

Your turn

Want a case study of your own?