WorkCase Study · Launchpad by Molt

6 processes mapped, first automation in production, in 4 weeks.

Launchpad is Molt's service that transforms how a business function operates by integrating AI into its workflows, decisions and dynamics. This is the real case of how Molt ran it on its own admin team, with the same rigor it applies with any external client.

Molt running Launchpad on its own operations
Client
Molt (admin ops)
Industry
AI consulting
Region
Colombia · Distributed
Services
Discovery · Design · Build · Adoption
Time-to-value
4 weeks
Key outcomes
6 processes · 1 automation live
TL;DR

The essentials

  • Launchpad transforms how a business function operates with AI: a structured diagnostic, effort-value prioritization, and a first automation in production in 4 weeks.
  • Molt ran it on its own admin team: same process, same rigor, same decision-maker commitment the service requires.
  • 6 admin processes mapped with AI potential, prioritized on an effort-value matrix.
  • A methodology toolkit (skills for time-tracking, session synthesis, discovery interviews) built during this engagement that now ships with every Launchpad client from day one.
The challenge

Validate the method on a real operation, no easy-client excuse

Launchpad is Molt's service that redesigns how a business function operates by integrating AI into its workflows: it diagnoses adoption, identifies inefficient processes, and builds tailored solutions (skills, MCPs, and automations). The north star isn't point automations: those are the entry point. The north star is a team that works differently.

Before offering it to external clients, Molt ran it on its own admin function with the same rigor it would demand of any client: Delta Loop, three options to the decision-maker at every milestone, the 80/20 rule (understand the problem before reaching for technology) from the first session. The admin team received no preferential treatment. The results are what they are.

The solution

Launchpad's four capabilities, applied in order

01 · Discovery
Discovery
Focused mapping of six admin processes. An AI-adoption survey across the team. Market research on how peer consultancies and similar firms use AI in their admin functions. A validation interview with the decision-maker.
02 · Diagnostic
Effort-value matrix
Cross-reference between the team's real pain points and market practices. A matrix of Quick Wins, strategic projects, minor improvements, and processes to avoid. The decision-maker received three implementation options at every step, never a single closed recommendation.
03 · Build
The lowest-exposure Quick Win
An admin routine the team ran manually every time (collecting information, logging it to a spreadsheet, notifying the team channel) fully automated via a skill connected to Google Sheets. The task stopped existing as human work. Three cost profiles evaluated before choosing.
04 · Adoption
Into production
The routine deployed into the actual production environment, with before/after measurement on the diagnostic dimensions. In parallel, the team built a toolkit of skills (for tracking time, synthesizing sessions, running discovery) that's now part of the service. Every Launchpad client gets it installed from day one.
Results

What changed

Four weeks in, Molt's admin team has clarity on its operations that didn't exist before. The team knows which six processes are candidates for AI transformation, the order to prioritize them by effort and value, and how Molt's AI maturity compares against peer consultancies. The first automation is installed in the production environment and runs without ongoing manual intervention.

As a byproduct: the methodology toolkit the team built while executing this case is now part of the service itself. Every client entering Launchpad receives a method tested on a real operation, not in a PowerPoint.

6
Admin processes mapped with AI potential
1
Automation running in production, zero manual intervention
4wks
From first discovery to automation in production
Key outcomesBeforeAfter
Admin processesInvisible, manual workExplicit map of 6 processes with AI potential
AI maturityAssessed with no external referenceBenchmarked against market actors
Quick winsIdentified informallyClear effort-value prioritization matrix
DecisionsMade without evaluated alternativesCompared cost, risk & benefit. The decision-maker chooses, not guesses
MethodologyTacit, undocumentedReusable toolkit, now ships with every Launchpad engagement
In their words

Nicolás Tobón was the decision-maker in this engagement, not the service designer evaluating his own work, but the internal client who had to approve every milestone, choose between three options at each step, and sign off that the automation was ready for production. This is what he said at close.

A clear diagnostic of where the company stands on AI adoption compared to the main players in the market, plus a set of action items to keep improving.
Nicolás Tobón, CEO, Molt
Next steps

Launchpad is available

Molt is extending Launchpad to its design function (diagnostic in progress). The service is available for external clients now: mature method, documented real case, toolkit included from day one.

FAQ

Frequently asked questions

How long will this take?
Four weeks from the first discovery session to a first automation installed in production, based on Molt's case applied to its own admin operations. Total scope depends on each engagement, but the first Quick Win in production is a milestone you'll see within four weeks.
How do I know the results will be actionable and not just a nice-looking report?
Launchpad never delivers a diagnostic without an effort-value matrix and at least three implementation options with differentiated costs. The decision-maker walks out of each session with decisions, not conclusions. In Molt's case, the diagnostic was cross-referenced with market research so action items had external reference, not just internal reading.
How does this integrate with the processes we already have?
Launchpad won't force your team to switch tools. We start by understanding what they use every day and how they use it; then we propose the simplest thing that solves the problem, built on top of what's already working. We only build something new when what's already there isn't enough. In Molt's case, the first automation worked on the same tools the admin team was already using daily, with no new tool to learn.
What do I tell someone who asks if Launchpad has external clients?
Launchpad has one documented real case (Molt's own admin function) with discovery, diagnostic, automation in production, and before/after measurement. The service is available for external clients now. If they ask for an external case: none published yet, but the method was proven on a real operation with the same rigor applied to any client.

Ready to change how your team actually operates with AI?

If your organization wants to transform how a function operates with AI, not just automate isolated tasks, Launchpad starts from diagnostic, before proposing any solution.

No commitment · 30-min intro call

Launchpad Case Study | Molt