Confidential proposal · May 2026

Build Morrison's AI operating layer, one real deployment at a time.

A long-term partnership to map the business, connect AI to approved systems, and deploy it against work that matters.

01 Map Discovery and workflow mapping.
02 Connect AI workbench and approved context.
03 Deploy Production systems for real bottlenecks.
Ongoing Operate Managed AI Ops keeps it useful.
BENALI
Morrison Cups - Long-Term AI Partnership
The foundation

AI works when the business is ready for it.

Most companies start with a tool and ask the business to adapt to it. We believe the order should be reversed: map how work moves, connect AI to approved context and systems, then deploy it into specific jobs with clear ownership.

For Morrison, that means AI should be grounded in the work Morrison already runs every day. It should understand the operating reality before it changes the operating rhythm.

The first job is to map Morrison's workflows and technology, then set up the foundation that lets the company adopt AI effectively instead of chasing disconnected tools.

The engagement path

Three phases, each with a distinct outcome.

Each phase creates the conditions for the next one. Phase 1 makes the work visible. Phase 2 gives AI approved access to the right context and systems. Phase 3 turns that foundation into deployed tools and workflows that do real operational work.

Phase 1
Map

Interview leadership and the people closest to the work. Map functions, handoffs, recurring workflows, and places where work piles up or depends on one person. Identify the first constraints worth deploying against.

Outcome A clear operating map and a prioritized list of AI opportunities grounded in Morrison's actual work.
$10K
Phase 2
Connect

Set up the AI workbench for Morrison's actual stack, confirmed in Phase 1. Connect approved systems, documents, data, and workflows. Set permissions so AI sees and does only what it should.

Outcome AI has the right context, tools, and boundaries to work inside Morrison's environment.
$15-25K
Phase 3
Deploy

Build production deployments against targeted bottlenecks. The pricing tool is the first Custom Deployment; Phase 1 turns that first win into an evidence-based roadmap for what should come next.

Outcome Specific systems ship into the business and start improving real workflows.
$10K+

Phase 3 options.

Morrison may need a workflow automated, a trusted company interface, a custom application, or some mix over time. These categories define what deployments we can land once the right bottlenecks are clear.

Workflow

Workflow Deployment

Best when a recurring workflow is slow, inconsistent, manual, or dependent on one person.

$10-35K
Agent

Company Agent

Best when employees need one trusted place to ask questions, retrieve context, and trigger approved workflows.

$35-60K
Custom

Custom Deployment

Best when the work needs a purpose-built application, internal tool, private AI setup, or deep integration.

$50K+

Ongoing: keep the system trusted.

Managed AI Ops keeps deployments reliable, connected data current, permissions reviewed, and future deployments improving after launch.

Essential

$2-4K / mo

Connector checks, light support, prompt and skill updates, and monthly usage review for shipped deployments.

Managed

$5-8K / mo

Everything in Essential plus workflow tuning, permission review, small quoting and workflow improvements, and adoption support.

Embedded

$10K+ / mo

Active roadmap, ongoing deployment iteration, custom monitoring, and priority support across Morrison's AI operating layer.

First deployment

The pricing tool is the first proof point.

The pricing tool is already live in the conversation. It touches quote volume, pricing logic, freight, HubSpot data, and production capacity. It is meaningful enough to prove the relationship, but bounded enough to ship before a company-wide AI program.

  1. Scope the pricing tool. Run the required technical scope and design session so the first build is clear before implementation starts.
  2. Build the first deployment. Ship a reliable quoting application with the maintenance path Morrison needs after launch.
  3. Use the build to learn the business. The project surfaces data realities, ownership patterns, and places where AI needs better context.
  4. Map the wider operation. Phase 1 finds the next bottlenecks without making guesses from the outside.
  5. Connect, then deploy again. Phase 2 gives AI approved access to Morrison's systems; future systems are chosen from mapped evidence.
Why Benali

AI transformation needs strategic judgment and practical builders.

The hard part is not buying AI. It is knowing where AI belongs, what should stay human, what needs better process first, and how to turn that judgment into systems people actually use.

Strategy

Business outcomes first

We connect AI decisions to operating goals, ownership, and measurable workflow improvement.

Transformation

Change that can land

We help leaders move from interest to adoption by shaping the path, the rollout, and the decisions around access.

AI depth

Modern AI architecture

We design AI systems around context, permissions, tools, prompts, skills, monitoring, and human review.

Builders

Custom systems, not slideware

We can map the opportunity, build the application, connect the environment, and stay close as it changes.

Fit

AI adoption works when leadership treats it like operating work.

The partnership works best when the people who own the business stay close to the map, the access decisions, and the adoption path.

Good fit

  • Operationally serious leadership team.
  • Clear recurring workflows and handoffs.
  • Enough urgency to implement, not just explore.
  • Willingness to give AI approved context and tool access.

What we need

  • Leadership in the room for operating decisions.
  • An honest picture of how work really happens.
  • Approved access to the systems and documents that matter.
  • Enough buy-in for the team to use what ships.

Recommended next steps.

First, run the required technical scope and design session for the pricing tool. Second, finalize the build scope and investment. Third, use that first deployment to inform Phase 1 Map and decide where the broader AI partnership should go next.

Khalil Benalioulhaj Work Architect · Benali
Nick Nance Technology Architect · Benali