Skip to content

Case study

One operator supervises around 100 accounts, versus around 5 managed manually.

For OCS, a marketing agency, I turned primarily oral social media processes into a supervised automation platform.

Client and sector
OCS, a marketing agency
Role
Lead Automation Engineer
Period
March 2025 – February 2026
Scope
Responsible for process framing and for designing, delivering and putting the automation platform into operation.

Verified result

Accounts supervised by one operator

Around 100

Manual capacity before

Around 5 accounts/operator

Orchestrated actions per day

Around 9,000

Method: around 20 operators is a projection based on around 5 accounts per operator toward the 100-account target. Around 9,000 actions per day is based on around 90 actions per account per day.

Approach

Business problem

OCS needed to scale a multi-account social media operation without multiplying manual execution.

  • Manual capacity capped at around 5 accounts per operator.
  • Oral processes that were hard to transfer, audit or improve.
  • A fragmented view of executions, blocks and interventions.

Response

I designed and delivered the platform that orchestrates the workflows while keeping sensitive decisions under human validation.

  • Covered setup, AI-assisted content, scheduling, publishing, warmup and semi-automated interactions.
  • Centralized supervision of social profiles, actions, statuses and errors.
  • Centralized action history and tracking information to understand blocks.

Before / after

  1. Capacity

    Before1 operator handled around 5 accounts

    After1 operator supervises around 100 accounts

  2. Process

    BeforePrimarily oral operating knowledge

    AfterStructured, executable workflows

  3. Supervision

    BeforeFragmented supervision and operator-dependent diagnosis

    AfterCentralized supervision and history

Let’s map your workflow, its volume, the tools or data involved, and its main failure mode.

Discuss a workflow to automate

Delivered work

Delivered system

  • Formalized execution steps, rules, exceptions and sensitive decisions.
  • Coordinated action planning at scale.
  • Centralized supervision of profiles, statuses and interventions.

Operating safeguards

  • Limits and pauses adapted to each profile.
  • Controlled recovery when an execution is blocked.
  • Action history and human validation for sensitive decisions.

What this case study demonstrates

  • Turn oral business processes into executable workflows.
  • Separate automated execution, supervision and human validation.
  • Make blocks understandable through action history and tracking information.

Discuss a workflow to automate

Let’s map your workflow, its volume, the tools or data involved, and its main failure mode.

Hugo Caulfield

Portrait of Hugo Caulfield