PRAXIS.
Operational Intelligence
PRAXIS.
Operational Intelligence · Est. MMXXVI
An operator-led software firm · Phoenix · Berlin · Est. 2026

Every off-the-shelf tool failed our operation. So we built our own.

Operators
who build.

Operational Intelligence is the system underneath your tools, the one your team currently carries in their heads. We build it, document it, and hand it to you.

Founder-led. Fixed-scope. One painful workflow at a time.

Live proof · drawn from production right nowCultivo·156 tables·12 modules·multi-state production today
Book a 20-min fit call See the sprint path One workflow. One working proof. No account team.
Cultivo · cannabis flagship · live stateOperational
active cycles7
canopy assignments196
in flowering3 cycles
in drying2 cycles
brain rows985
One Praxis-owned product, running production today. The same patterns ship in every client engagement.
FIG. 01 · THE OPERATING ATOM · DRAWN FROM LIVE DATA

An orbital schematic of the firm. Every entity, every engagement, every table, drawn from the live production infrastructure.

Most firms describe themselves with words. We render the actual structure. The atom below is generated from the same context database that runs Cultivo in production. The center is the mark. The rings are operators, products, methodology, engagements, and the tools layer above. Drag to rotate. Hover any node to inspect.

“Most firms have an org chart. We have an operating atom, drawn from the same database that runs production. If a relationship breaks in the data, the line breaks here.”
Open full atom
§ Engagement velocity

Five engagements, at five points in motion.

Each particle below is one anonymized engagement, plotted at its current position on the methodology pipeline. As an engagement moves through Discovery Sprint into Build Sprint into operating retainer, its dot drifts right. Whoever lands the open slot becomes a sixth particle, starting at the left.

§ Pipeline · 2026 · anonymized5 active · 1 open
drifts left → right · auto-updated
Discovery Sprint
Five business days
One workflow · working proof
Build Sprint
4 to 6 weeks
Production workflow · fixed scope
Operating Retainer
Monthly cadence
Continuous improvement after launch
ENGAGEMENT 01Build · week 6
ENGAGEMENT 02Build · pre-week 1
ENGAGEMENT 03Sprint · week 1
ENGAGEMENT 04Qualify
ENGAGEMENT 05Qualify
02 in Build01 in Sprint02 in Qualify01 OPEN
Qualify Sprint Build
0%
of companies fail to see ROI from AI investment
0%
of AI budget goes to tools, not the operations underneath
0+
operational modules shipped in Cultivo, in production today
0 YRS
operating in production before we sold our first engagement
Operators we build forHospitality consultingReal estate platformsField-services opsMulti-state retailCannabis cultivation
§ The pain

Four ways operations break before software ever helps.

PAIN 01
I can't leave.
The operation breaks without you. Decisions don't get made. You haven't taken a real vacation in two years.
Trapped operator
PAIN 02
The data is everywhere.
Six tools. Three spreadsheets. Tribal knowledge in five heads. No source of truth, no two reports agree.
Scattered systems
PAIN 03
AI didn't fix it.
You bought the tools. The vendors promised the layer underneath. The reports are still wrong.
Stalled investment
PAIN 04
The next hire won't help.
A fractional CTO advises. A dev shop bills hours. Neither builds the operational layer underneath the tools.
Wrong shape of help
§ Thesis · the seven-percent layer

93% goes
to the tools.
7% goes to
the layer.

93%
7%

Most AI and software investments fail because the operation underneath was never documented. The data is scattered. The logic lives in people’s heads. The tools sit on top of that gap and produce reports nobody trusts.

We know because we lived it. We ran a multi-state operation, watched every off-the-shelf tool fail us, and built the infrastructure ourselves. Now we build it for other operators.

Cultivo is the proof: 156 tables, 12 modules, running in multi-state production since 2024. The hierarchy problem and the tribal-knowledge problem are the same problem. We felt it, then we built the layer underneath.

§ Capacity
06.
Engagements per year
Founder-led. No account team. No handoff. No offshore. When we are full, we refer.
§ Disqualification

Who we are not.

01
Not a dev shop.
We do not bill hours.
02
Not a fractional CTO.
We ship code, we do not just advise.
03
Not an AI consultancy.
We do not deliver slide decks.
04
Not a venture studio.
We do not take equity for services.

We are operators who build software for other operators. Discovery Sprint first, then fixed-scope implementation when the workflow is worth rebuilding.

01 open sprint slot · Discovery Sprint · five business days
§ How we work

Three engagement shapes. Pick the one that fits the problem.

SHAPE 01
Discovery Sprint.
Five business days · one workflow, inspected and proven.
We pick one workflow that is costing time, revenue, or owner attention, map how it actually runs, and prove the fix with a working proof or demo. You leave with a roadmap and a clear build recommendation.
Fixed-scope first step · priced after fit
SHAPE 02
Build Sprint.
Four to six weeks · one workflow shipped to production.
The proof becomes an operator-grade workflow: connected to real data, reviewed by humans where it matters, documented for the team, and supported through launch.
Fixed-scope implementation · no hourly billing
SHAPE 03
Operating Retainer.
Monthly cadence · continuous improvement after launch.
For teams that want Praxis to keep improving the operating system month by month: workflow automation, reporting, AI assistant maintenance, and new decision-support surfaces.
Retainer available after implementation

From the Praxis operator’s manual · 2026
§ Proof of work

Real systems. Real operators. Real numbers.

CULTIVO.Cannabis operational platform · 2024 → 2026 · in production
From three spreadsheets and one person’s head to a multi-state operational platform. Cultivation, inventory, sales, and finance, on one source of truth.
156 canonical tables. 12 operational modules. Daily users across cultivation, post-production, sales, and finance. Running in multi-state production. Praxis-owned product, formerly CultOps. The flagship evidence of how we build, and the foundation of every methodology we apply to client work.
0
canonical tables in production
0+
operational modules shipped
99.94%
uptime over 30 days
cultivo.ag →· Multi-state production
ASPIRE.Hospitality consulting · 2026 · active engagement
From a website that read like a 2014 brochure to a clear positioning surface and a diagnosed roadmap. Website rebuild shipped. Discovery Sprint delivered.
Praxis rebuilt poweredbyaspire.com, then ran a structured Sprint surfacing where the operation breaks and what to build next. Findings delivered at aspire.gopraxis.ai. Detailed findings available under NDA on request.
P1
website rebuild · shipped 2026
Sprint
complete · findings delivered
P2
build proposal · in scoping
poweredbyaspire.com →Sprint delivered · aspire.gopraxis.ai
§ The firm in operation · right now

A console, not a brochure.

Production state, capacity, integrity. Drawn from the same context database that runs Cultivo in production. If we are operating, this console is moving.

§ PRAXIS · OPERATIONS CONSOLE · v1.0UTCSTATIC · refresh on next deploy
§ Production system · Cultivomulti-state
tables156modules12uptime · 30d99.94 %scopecult · inv · sales · finin productionmultiple states
§ Capacity boardAnnual · 2026
06/ yr · max
Engagements per year
01 open
Available now
By stage · annual pipeline
02in Build
01in Sprint
02in Qualify
01OPEN
06Open sprint slot · Discovery SprintBook fit call
§ Operators on dutyFounder-led · no account team
Justin Morrow.
AI engineering · product architecture · technical lead
JM-001ACTIVE
Greg Dunaway.
Client · partnership · engagement structuring
GD-001ACTIVE
§ IntegrityAll systems
schema lineageOKverified at deploy
RLS policies4 layersOKverified at deploy
context dbliveOKverified at deploy
backupsdailyOKlast 06:00 UTC
env / secretsconfiguredOKverified at deploy
deploy gatenpm buildOKlast build at deploy
drawn from praxis.brain · console v1.0STATIC · live wire pending
§ People

Two operators. No account team. No handoff.

JM.
Justin Morrow.
Co-Founder · Managing Partner · AI engineering & product architecture
I ran cannabis operations and watched every off-the-shelf tool fail. So I built the layer underneath it: 156 tables, 12 modules, lineage contracts, the whole production system. Now I lead the technical build on every Praxis engagement, and apply the same patterns we proved in our own operation.
· justin@gopraxis.ai
By intake or email
GD.
Greg Dunaway.
Co-Founder · Managing Partner · Client & partnership strategy
Fifteen years operating, four companies founded, one through exit. I have lived every version of this conversation from the operator side. At Praxis I run the relationships, structure the engagements, and stand next to the work in front of the client when the engagement calls for it.
· greg@gopraxis.ai
By intake or email
§ Intake

Send one workflow worth inspecting.

Justin reads every submission. If the workflow is a fit for a Discovery Sprint, you will hear back with a clear next step. If it is not, you will still hear back.

Already know we should talk? Book a 20-minute fit call with Justin →

Discovery Sprint first. One workflow, one working proof.
§ One last thing

Who should not file.

If you are looking for hourly engineers, a fractional CTO, or a slide deck, close the tab. We are not the firm.

If you have one workflow that keeps leaking time, revenue, or attention, send it in. Justin will read it and tell you whether it is a fit for a Discovery Sprint. Either way, you hear back.

Justin & Greg.
Co-Founders · Managing Partners