Cross-Surface Workflow

The clearest way to understand BlazeXL is to follow one piece of logic across its full lifecycle.

This is the canonical BlazeXL workflow:

Draft in a familiar surface -> publish as a governed Block -> execute across Excel, apps, APIs, and agents -> review in audit

Step 1: Start Where Logic Already Exists

Most teams do not begin in a platform-native editor. They begin in:

• Excel
• Python scripts
• notebooks
• AI-assisted drafting flows

BlazeXL is designed to meet teams there.

Example starting points

• a pricing analyst prototypes logic in Excel using the add-in
• an operations team drafts reconciliation logic with AI assistance
• a data team turns an existing Python function into reusable logic

At this stage, the goal is speed and iteration.

Step 2: Turn the Logic into a Blaze Block

Once the logic matters to the business, it should move out of ad hoc execution and into a governed Block.

A Blaze Block gives the logic:

• versioning
• permissions
• auditability
• reuse across surfaces

This is the point where spreadsheet logic or AI-generated logic becomes governed organizational logic.

Step 3: Use the Same Block Across Surfaces

After publication, the same logic can be executed from multiple surfaces without duplication.

Excel

Use the add-in to call the approved Block from a worksheet.

=BLAZEXL.PY("pricing.calculate_quote", A2, B2, C2)

Blaze Apps

Wrap the Block in a structured interface for business users.

Internal Service or API Client

Import the approved logic into a session and execute it through the SDK.

import { BlazeClient } from '@blazexl/client-node';

const client = new BlazeClient({
  token: process.env.BLAZEXL_TOKEN,
  baseUrl: process.env.BLAZEXL_URL,
});

await client.importModule({
  shareId: process.env.BLAZEXL_PRICING_SHARE_ID!,
  name: "pricing",
});

const quote = await client.execute("pricing.calculate_quote(args['arg1'])", {
  arg1: {
    region: "APAC",
    segment: "enterprise",
    amount: 125000,
  },
});

Agent

An agent can orchestrate work while BlazeXL executes the governed business logic behind the tool call.

The important point is that the agent does not need to invent the business logic at runtime. It can call approved logic under an attributed identity.

Step 4: Review What Happened

Every execution surface feeds back into the same governance model.

Teams can review:

• who or what ran the logic
• which version was used
• which surface invoked it
• whether the actor was human or agent
• execution status and metadata

This is what makes BlazeXL more than a compute platform. It is an execution standard with reviewability built in.

Example End-to-End Pattern

Spreadsheet-led workflow

1. Analyst drafts pricing logic in Excel
2. Logic is reviewed and published as pricing.calculate_quote
3. Workbook enters strict mode and calls the published Block
4. Internal service also imports the same Block through the SDK
5. Procurement agent calls that Block through a dedicated agent identity
6. Audit shows human, API, and agent usage of the same approved logic

Why This Matters

Without BlazeXL, each surface tends to grow its own copy of the logic:

• spreadsheet formulas
• local scripts
• backend services
• AI-generated code snippets

That creates drift, review gaps, and inconsistent decisions.

BlazeXL gives teams one governed logic layer that every surface can use.

Related Guides

• Excel Add-in Installation
• Blaze Sheets
• Governed API and Agent Access
• Grounded AI Analyst