Is platform logging enough?

Not if the gateway hop, the broker hop, and the tool hop live in three places with no shared identity. We will say when you already have enough.

Do you skip human review when confidence is high?

High-confidence reads can auto-complete. Write-back on money, identity, and policy still hits a person until the evals say otherwise.

AI observability

If you cannot see the hop, it is not in production.

Traces, identity, and cost on one trail. A person on the exceptions. A log dump is not observability.

Get the written assessment Agent Broker

Founded 2022 · Founder in the Salesforce MVP Hall of Fame · Dubai HQ · Kochi studio

Trail

Gateway, broker, agent, tool. One trail.

The same path the architecture uses. Dark hops are findings.

01
Gateway
Who called, which policy, which model, how many tokens. This is the edge.
02
Broker
Which skill was chosen. Automation, guided, or agentic.
03
Agent
Which topic, which action, which card. Write-back still fails closed.
04
Tool
Which MCP server. Success, retry, or dead letter.

Public observability path for agent and integration flows. Every hop carries identity. Cost sits on the same trail as the trace. Exceptions land where a person can see them.

Mermaid source

flowchart LR
  G[Flex Gateway] --> B[Broker]
  B --> A[Agent]
  A --> T[MCP tool]
  G -. identity and cost .-> O[Trail]
  B --> O
  A --> O
  T --> E[Exception queue]

Log or trail

A log dump, and a trail.

A log dump

One trail

Lines per service, no shared ID: One trace across gateway, broker, agent and tool
Cost on a separate cloud bill: Tokens and tool calls tagged per caller
Failures in an inbox: Dead letters in a watched queue
Nobody named: An owner named in the operating model

Exceptions

Retries, dead letters, and a person.

Extraction fails. A tool call fails. The hop times out. Someone has to see it.

What it catches

Failures we see.

Example
Platform lead at a Dubai bank
Today: An agent answer was wrong and nobody could say which knowledge it used.
After: The trace shows the retrieval, the topic and the version that answered.
Example
Finance controller at a US retailer
Today: AI spend jumped one month with no owner.
After: Usage is tagged by desk, so the spike has a name.
Example
Integration lead at a Riyadh manufacturer
Today: Document extraction fails quietly at night.
After: Failures land in a queue a person clears each morning.

Examples of the work, not case studies. Roles and cities are typical of the desks we see.

Cost

Token and tool usage tagged per caller.

Finance should see a process, not a cloud bill.

Questions

What we actually say.

Is platform logging enough?: Not if the gateway hop, the broker hop, and the tool hop live in three places with no shared identity. We will say when you already have enough.
Do you skip human review when confidence is high?: High-confidence reads can auto-complete. Write-back on money, identity, and policy still hits a person until the evals say otherwise.

The brief · one email

Show us one failed hop.

We will tell you whether you can see it, and who owns it. The assessment is free: a short read of the org and a written note. If you go ahead, the first build week is a fixed AED/SAR 14,900 (about USD 4,050), credited in full when you sign the build within 60 days. The honest no: if the work is a software factory with a named delivery date, we are the wrong partner.