Capability

Agentic solution development.

Business systems are moving from applications that record and route to agentic surfaces that reason and act. That is a paradigm shift, not a feature, and we are driving it. This capability is how: the ability to architect, build, integrate and run complex agentic solutions for many clients at once, as managed solutions. The paradigm is on its own page; this page is the engineering.

Start No-Risk Discovery The paradigm, screen by screen
What it is made of

Six parts, and the client already owns one of them.

The applications of record stay where they are. Everything else is what we build.

The control plane.

Sits above the applications, not inside one. It holds the operating model (what good looks like, the rules, the gates), the skills that run each workflow, the run log and the conversation.

The skin: one surface.

A custom set of screens over the client’s applications, every screen carrying a conversation pane that knows the page, so people talk or click on the same screen. The same brain answers in the chat tools they already use.

Agents with gates.

An agent for each job: it reads across the systems, drafts the next step and runs the obvious ones. Anything that costs money, carries risk or needs taste waits for a person’s one-digit answer.

Routines and workflows.

The scheduled runs (nightly reads, daily observe and diagnose, the weekly slate, the monthly review) and the event-driven ones (a lead arrives, a reply lands, a threshold is crossed), each with a scope, a gate and a failure behaviour.

The data and audit layer.

One performance table across every source, and one row per action in the run log: who asked, from which surface, who approved, what the API returned, what it cost. The audit trail and the source the reports read.

The integration layer.

Gateway functions to the client’s own platforms through their APIs, credentials in an encrypted store no person or prompt can read, every write read back. Platform agnostic: any vendor in a category whose APIs carry the use case.

The architectures

Six patterns we build every solution from.

None of them is exotic. Together they are what makes an agent safe enough to sell and cheap enough to run.

  • One surface over many systems. The skin and its pane are the face; the applications are the data plane. A person works one set of screens designed for the job, not one vendor screen per application.
  • Event-driven routines. Work starts when something happens or when the clock says so, never when someone remembers. Every routine has a scope, a cadence, a gate and a defined behaviour when it fails.
  • Human gates. Every dollar has a gate a person controls; caps are enforced in code below the gate. A gate raised on a screen can be answered from a phone, and the answer lands on the same record.
  • Read before write, and read back. A workflow reads the current state in the same turn it writes, writes once, then reads the result back and compares it. A read-only probe never proves a write path.
  • The ledger of runs. One row per run and per action, with its cost. It is the audit trail, the input to the all-in cost of every outcome, and the evidence the system learns from week to week.
  • Platform-agnostic adapters. The control plane never calls a vendor directly. Gateway functions own each platform’s API, so a client’s CRM, marketing or finance system can be any vendor with the APIs the use case needs.

Four layers, top to bottom.

  1. The skin with the pane

    The custom screens and, on every one, the conversation pane that sends the page context with each message and refreshes when the plane reports a change.

  2. The control plane

    The operating model, the skills that run each workflow, the gates (thresholds and approvers per action), the conversation handling for every surface. Credentials live here, encrypted.

  3. The applications’ APIs

    The client’s CRM, marketing, finance, help desk, inventory, analytics and the systems around them, called by the plane’s gateway functions; every write read back.

  4. The run log

    One row per action: who asked, from which surface and screen, the gate and who passed it, the call and its read-back, the cost.

What it does

Plan, propose, produce, publish, measure. A person at each gate.

The verbs are the same in every function we have built for. The example is our own marketing, because it runs today and you can watch it.

Plan.

Themes scored against search demand, competitors and the pipeline every month; programs with a hypothesis, a window and a kill rule.

Propose.

The week’s slate, the budget and the next move, read overnight from the numbers and posted with its reasons. Three decisions, one digit each.

Produce.

Posts, videos, emails, social copy and ads made to the plan, each to its own gate, each with its cover, its tags and its cost on the row.

Publish.

To the CMS, the video channel, the ad account and the email list on schedule, never before the page it links to answers, every URL checked in the same run.

Measure.

One performance table from ads, analytics, search, CRM and email; qualified leads sent back to the ad platform; the all-in cost landed on every piece.

The routines behind them, as they run today.

  • A daily read of every source into one performance table, and a daily marketing read against it
  • Daily observe, diagnose and signal on paid search, under a monthly envelope enforced in code
  • A weekly slate of pieces and announcements, the newsletter and the weekly summary
  • A weekly search-health run: backlinks, disavow, site audit, crawl
  • A monthly theme and program review, and the all-in cost run
  • A nightly self-check that tests every connection and restores what it finds broken
The effort it saves

Two axes: building the solution, and running the function.

Measured on the solution we run for ourselves, AI Marketing OS, by industry-typical task times for the work as specified. These are the numbers we show a client, with the method.

Building it.

The first twenty days of the build (1 to 20 September 2026) delivered 4,442 hours of work by conventional task times: 27.8 person-months of a conventional build, in 20 calendar days. That is 222 hours of output a day, 9.3 times one person working around the clock. A conventional team of the same size would still be in discovery.

hours of work delivered in 20 days
4,442
person-months of conventional build
27.8
hours of output per calendar day
222

Running it.

As designed, the routines deliver about 1,380 hours of work a month for 43.5 hours of a person’s supervision: 8.6 full-time jobs replaced, and the management those jobs needed cut by 86 percent, from 2.27 to 0.31 full-time equivalents. The tooling costs 400 dollars a month before media.

hours of work a month, as designed
1,382
full-time jobs replaced
8.6
less management needed
86%
By job function, hours of work a month.
FunctionHours a month
Blog, twenty posts a week325
Video, ten videos a week624
Search health and crawl, weekly200
Measurement, daily75
Paid search, daily38
Newsletter and weekly summary37
Programs and themes24
Influence, social, self-check, replies59

Each row is units a month times industry-typical hours per unit, never a measurement of our staff. Published trials of AI tools on knowledge work measure 0.84 to 1.67 times one person’s output; a governed system of agents is a different thing from a tool in a person’s hands, which is the point.

The labour that goes is the coordination, the re-keying and the management of both. The deciding stays with a person.

For many clients

One control plane, one instance per client, run as a managed service.

Building one agentic solution is engineering. Running them for many clients, safely, is the capability.

A solution instance per client.

Your records, your credentials in your encrypted store, your gates and your approvers, your run log. Nothing shared with another client’s instance.

A shared control plane.

One codebase and one set of versioned skills run every instance. A fix or an improvement reaches every client as a release, not as a project.

Versioned skills and change control.

Every skill ships with its tests; a release checks the production surface before and after the change and fails on any lost field. Nothing reaches a client untested.

Monitoring and self-healing.

Connections and routines are tested daily. A regression is restored to the code of record and read back in the same run; a disconnected connector is reconnected. You hear about it as a line in a report, not as an outage.

The advisory layer.

A system runs what it is given. A senior person reviews the plan, the themes, the targets and the results with you, so the inputs stay sound and the outputs stay worth paying for.

A subscription, in tiers.

The software, the integration, the operations and the advice for one monthly price, in tiers per solution set by monthly usage. The estimate for setting it up on your platforms is guaranteed.

What a managed solution includes
Related

Custom software development.

The engineering practice behind the agents, the skins and the adapters.

Custom software

Managed technical operations.

How the running of a solution is staffed and measured.

Managed operations
Go deeper

Keep going on agentic solutions.

Related articles.

A single card holding five horizontal cost bars of different lengths, one drawn in purple and the rest in pale grey, above a green total rule with a navy and a green block beneath it.
Articles · Sep 28, 2026 · 6 min

What It Costs to Run a Marketing Function as a System: Our All-In Numbers.

Most companies buy marketing as headcount or a retainer and judge it on activity. We run ours as a governed system and know its all-in cost per piece. Here are the numbers, what they leave out, and how to read them against a hire or an agency.
  1. Articles · Sep 6, 2026 · 6 minNobody Is Governing the Agents: What Yours Can Reach, and Who Decided.
  2. Articles · Sep 6, 2026 · 6 minThe Zoho MCP Server Opens Your Estate to Outside Agents. Set the Permission Model First.
  3. Articles · Sep 28, 2026 · 6 minCRM Adoption Is Not the Problem. What Breaks When the Foundation Is Thin.

Related pages.

TechnologyAgent development.Agents that read and act across the applications a business already runs, with a person approving what matters. TechnologyAI-driven software development.Agents write, test and review code against written rules; a person gates every release and every change leaves a record. Managed solutionLight Manufacturing, eCommerce & 3PL.From purchase order to margin by channel, one plant floor in chat. Managed solutionAI Marketing.Marketing run as one governed system, from theme to lead.

Related videos.

All articles on agentic solutionsAll videos in AI Marketing
Where we’re not the right answer

This is for solutions we run. For hours of work, see the services.

We take on both today, and we will say which one you are buying.

  • You want configuration or integration work by the hour on a platform you run yourself. That is services work, available today on our services pages. It is not this capability, and we will not dress it up as one.
  • Your work does not cross systems and nobody needs to approve anything. A configured application will do, and we will say so.
  • You want an agent with no gate. We build the gates in. If that is not what you want, we are not the right partner.
How we start

Most of our best clients come to us with a feeling, not a plan.

"Something isn't working." "We're outgrowing our tools." "We're afraid to make the wrong move." No-Risk Discovery is a short, practical conversation that gets you clarity before you commit to anything big. We'll tell you if we're a fit. If we're not, we'll tell you that too.

Start No-Risk Discovery See the solutionsWhy We're Different