Skip to content

Evidence-led AI operations

Find the leverage. Prove the value.

Velomensa helps Quebec SMEs reduce repetitive follow-ups, manual handoffs, and operational errors. Start by identifying one useful improvement, then measure what changes. Available in French and English across Canada.

  • Production-minded
  • Human-controlled
  • Measured evidence
OUR METHODVM-01

Evidence guides the next step.

Read the steps
  1. Diagnose: establish the baseline and scope.
  2. Deploy: implement with human controls.
  3. Measure: observe results and costs.
  4. Decide: expand, adjust or stop according to the evidence. Expansion starts a new scoped cycle.

An illustration of the method, not live system status.

Hidden operational drag

Slow recurring work quietly taxes the whole company.

Copy-paste work, fragile handoffs, delayed follow-ups, and knowledge held by one person create costs that rarely appear on a dashboard. We make that friction visible before proposing technology.

01Time

Repeated coordination and manual updates

02Income

Slow response, missed follow-up, and leakage

03Reliability

Errors, exceptions, and inconsistent execution

04Knowledge

Critical process context trapped in people’s heads

A deliberate operating loop

Diagnose. Implement. Measure. Expand.

01

Diagnose

Map the work, quantify friction, and define the boundary.

Boundary: baseline + priority
02

Implement

Build a production system with clear ownership and controls.

Deliverable: production workflow + controls
03

Measure

Compare the baseline with observed operating evidence.

Evidence: 30-day observed result
04

Expand

Use proven results to choose the next highest-value move.

Decision: expand, optimize, or stop

Two ways to begin

Start with clarity or start with one proof.

01 · Diagnosis first

Automation Leverage Map

Map the company. See where time, income, errors, and fragile knowledge create the greatest opportunity. Leave with a prioritized roadmap.

View details
02 · Prove first

Value Return Sprint

Improve one process. Put a defined automation into production, measure 30-day results, and use the evidence to choose what comes next.

View details

Automation Leverage Map

See where automation can create the most business value.

We examine departments, processes, bottlenecks, income leakage, time demand, SOP readiness, dependencies, and risks—then score and sequence the opportunities into a practical roadmap.

View details
Which opportunity should we tackle first?

Illustrative example · Final decisions use your data.

Read the explanation

Illustrative matrix. First is the highest-impact opportunity with the easiest implementation, followed by Next. Validate, Prepare, and Defer identify lower-readiness or lower-impact options.

  1. First: High impact · easy implementation · high confidence Illustrative position, not a measured score.
  2. Next: Moderate impact · moderate ease · medium confidence Illustrative position, not a measured score.
  3. Validate: Lower impact · lower ease · medium confidence Illustrative position, not a measured score.
  4. Prepare: Relatively high impact · harder implementation · medium confidence Illustrative position, not a measured score.
  5. Defer: Low impact · difficult implementation · low confidence Illustrative position, not a measured score.
What changes between baseline and day 30?
  1. DAY 00Baseline + boundary

    Verify the starting condition and the result that will count.

  2. LAUNCHProduction + controls

    Release the bounded workflow with an owner and review gates.

  3. DAYS 01–29Observe operations

    Record real volume, exceptions, time returned, and reliability.

  4. DAY 30Review the evidence

    Compare the observed period with the baseline and choose the next move.

DECISIONExpand, optimize, or stop

Value Return Sprint

Put one important automation into production—and measure what changes.

Define one process, establish the baseline, build the production system, retain the right human controls, and observe results for 30 days.

View details

Business Value Ledger

Evidence that separates what happened from what is projected.

The ledger records the original baseline, observed results, time returned, costs reduced or avoided, reliability improvements, assumptions, and annualized projections.

What did we know, observe, and project at each stage?

Illustrative method · Not a client result · Every value retains its source, period, method, and reviewer.

Read the explanation

Illustrative lifecycle: a 120-hour baseline is recorded on day zero, a named estimate is retained as an assumption, 46 hours returned are observed during 30 days, and 552 hours are annualized only as a projection based on comparable monthly volume.

  1. Baseline: 120 h / 30 days DAY 00 · The verified starting condition before production. Source: workflow review · Current on day 00
  2. Estimated: Named assumption BEFORE LAUNCH · A hypothesis recorded explicitly and never presented as fact. Source: scoped estimate · Not observed
  3. Observed: 46 h returned LIVE RUN · What changed during 30 measured operating days. Source: operating ledger · Updated day 30
  4. Annualized: 552 h projected REVIEW · A transparent projection, not an observed annual result. Method: 46 h × 12 comparable months

Human control by design

Automation should make responsibility clearer—not remove it.

Sensitive decisions, exceptions, approvals, and client communications remain assigned to people. Every production workflow has an owner, an escalation path, and an auditable change history.

APPROVALEXCEPTIONOWNERAUDIT

Operate and improve

A long-term operating partner when the first win earns the next phase.

We can monitor service health, resolve incidents, review exceptions, improve workflows, and keep the value ledger current—without forcing an all-at-once transformation.

View details
Theodor Badele, founder of Velomensa

Founder-led accountability

AI transformation should earn the right to expand.

Velomensa was founded by Theodor Badele on a practical idea: begin where the business has leverage, build around real operations, and let evidence guide the next phase.

Theodor Badele FOUNDER / CEOView details

Community initiative

Separate from paid client services

Help the next small local business get online.

When capacity allows, Velomensa selects a small local business and builds a focused website without charging for the work. Voluntary community support helps make the next project possible.

This initiative is separate from Velomensa’s paid client services. Support is completely optional.

Explore the initiative

Choose the right starting point

Automation Leverage Map

Starting point
You are unsure which improvements to prioritize.
Deliverable
A process map and a prioritized roadmap.
Your participation
Describe operations and validate baseline information.
Boundaries
Diagnosis does not include implementing every improvement.
Discuss it on a discovery call ↗

Value Return Sprint

Starting point
You already have one process in mind.
Deliverable
A bounded automation and a measurement register.
Your participation
Provide a process owner, approve the rules, and test the workflow.
Boundaries
One agreed process; integrations and exclusions are defined before work starts.
Discuss it on a discovery call ↗

The 30 days refer to observation after launch, not a universal delivery timeline or a savings guarantee.

Illustrative sample · No client results

What does a measurable improvement look like?

Before: quotes and follow-ups are tracked manually. Proposal: identify pending quotes and prepare follow-ups for review. A person approves each message; a reply or refusal stops follow-up.

A follow-up, with your approval.

Illustrative method · No client results

Read the explanation

An eligible quote prompts a draft. A person approves every message. A reply or refusal stops follow-ups; without approval, nothing is sent.

  1. Identify: Pending quote Check eligibility and follow-up rules.
  2. Prepare: Follow-up draft Assemble the context, without sending.
  3. Get approval: A person decides Approve, revise or reject. No approval means no send.
  4. Respect the reply: Reply or refusal → stop After an approved send, a reply or refusal ends follow-up.
Baseline fields
Eligible quotes, follow-up minutes, missed follow-ups, and responses.
Costs to include
Setup, tools, human review, maintenance, and exceptions. Measure the values; do not invent them.
Comparison method
Compare similar periods and volumes. Separate observations from assumptions; document changes and limitations.
Explore practical guides ↗

Talk through one process

Start with a focused discovery call.

Bring one real process, its point of friction, and the outcome that would matter. Booking runs through Velomensa’s self-hosted Cal.diy instance.

Book a call

FAQ

Questions before the first move.

01What if we already know which process to automate?

Start with a Value Return Sprint. We validate the boundaries, baseline, controls, and measurement plan before production.

02Do we need a full company transformation?

No. Work proceeds in deliberate phases, and every phase must justify the next.

03How is value calculated?

Every ledger names the baseline, source, observed period, assumptions, exclusions, and projection method.

04Will we need to replace our current tools?

Usually not. We first improve and connect the systems your team already relies on.

See all questions
01 · CAL.DIY

Book a discovery call

A focused 30-minute conversation with the founder to understand the process, friction, and desired outcome.

Attendees
Process owner and decision-maker, when possible
Preparation
Bring one real example; no confidential data required
Timezone
Shown for your location
Open scheduling

Scheduling unavailable? Send a request.

Scheduling opens on Velomensa’s self-hosted Cal.diy instance. You can reschedule or cancel from the confirmation email.
02 · DIRECT MESSAGE

Or briefly describe the work.

Explain the process and what is slowing you down. We normally reply within two business days.

Next step

Bring the process that keeps slowing the week down.

A discovery call is a focused conversation about the work, the friction, and whether there is a useful next step.

Book a call