All work

Client work through InsideBoard

Automating coaching and improvement programmes

Managers and trainers submit one form. Feedback, scores and tracking reach the right profile without anyone copying them across.

I work with clients to understand how their programmes run, challenge requirements and turn them into InsideBoard configuration and automation.

Forward Deployed Engineer, InsideBoard AI · July 2026 to present · Enterprise SaaS, people development

Workflow overview
  1. A form is submitted

    A manager or trainer submits feedback for a person or team.

  2. The workflow applies the rules

    The workflow identifies the recipient, computes the relevant result and prepares the profile update.

  3. The profile and tracking are updated

    Feedback, scores and programme tracking are written to the relevant views.

Overview of the documented coaching workflows. The exact scoring and visibility rules depend on the programme.

The problem

Large organisations use InsideBoard to run coaching and improvement programmes: new-hire training, field coaching, team action plans. Managers and trainers fill in forms, but a form on its own changes nothing. Each answer has to reach the right person's profile, update the right score and KPI, and respect who is allowed to see what, across hundreds of people and teams.

The outcome

Feedback now reaches each coached person's profile, checked across all 123 people in field coaching. Training results are gathered in one view for trainers, and site diagnostic scores are calculated in the platform. The work covers customer-care training, field coaching and improvement plans, and rollout continues workflow by workflow.

My contribution

I lead client discussions to clarify who uses the programme, what should happen after each action and how success is measured. I challenge requests against the platform’s capabilities, resolve scoring and visibility rules, and turn those decisions into a solution design. I choose native forms, action plans and KPI configuration where they are sufficient, adding n8n and API integrations where needed. I then implement and test the solution with the InsideBoard team, refining it through delivery reviews.

The specifics

  • Forms connected to profiles, scores and KPIs through automation and native configuration
  • Field coaching: feedback appears on each coached person's profile, checked across all 123 people in scope
  • New-hire training: daily feedback, quiz score and certification decision gathered in one view per trainee, with clearly labelled AI readiness briefs for trainers
  • Improvement plans: site diagnostic scores computed and checked in the platform, with manager-only visibility configured

How it works

  • Workflows rebuild each profile section in full on every run, so late or repeated submissions never overwrite each other
  • A scoring bug in a live programme was traced to its root cause, fixed and backfilled for all 17 affected trainees, after explicit sign-off
  • Daily KPI jobs measure the delay between a field visit and its feedback, and count feedbacks per manager once per person and visit
  • The next form is pre-filled with the latest feedback, looked up by visit date, so managers start from the current state
  • AI briefs are optional by design: if the model call fails, the section is still written without them

Safeguards and quality

  • Client discussions become business rules, explicit decisions and a written solution brief that guides implementation
  • Each programme has its own repository with one status record, dated evidence and a gate that blocks changes to live systems until they are approved
  • Reusable platform knowledge is collected into a shared knowledge base, and boundary checks stop one client's data from reaching another client's project
  • Native configuration first: automation is added only where the platform cannot do the job itself

Stack

  • n8n
  • InsideBoard API
  • InsideBoard Forms
  • Paperform
  • LiteLLM
  • Fathom
  • Claude Code
  • GitLab CI

Have a process like this one?

aityahia@meetmarwane.com

More work

© 2026 Marwane Ait Yahia · Paris, France

Meet Marwane · Automation and AI agents for your team.

On assignment at InsideBoard, available for selected freelance projects in parallel.