See how fluent your company is in AI.
A certification ladder, a live workshop every month, and a dashboard your board can read. For companies of 100 to 1,000 people.
Run by a practitioner with a finance and technology background, not a presenter. Every pilot ends with a debrief and a written reading of the numbers, whatever they show.
What the numbers say: Operations is certified but at 15% Level 2 and 52% attendance: a conversation, not another course. The dashboard tells you where to look, not what to conclude.
Your people have tried AI. Most stopped at the first prompt.
Single-touch training plateaus below 25% adoption at 90 days; multi-touch cohort programmes sustain 60 to 75%. That gap is why Fluentia runs monthly, not once. Coursera Enterprise and LinkedIn Learning patterns, cited by Alice Labs, 2026.
Dabbling, not delivering
A few enthusiasts, many sceptics, and no shared standard for what good looks like.
Tools move monthly
What your team learned in spring is out of date by autumn. Nobody owns keeping up.
No number for the board
"We're doing AI training" is not an answer to "how capable are we, and is it improving?"
Fluentia is the ladder you can watch climb.
Three levels, earned by doing
Level 1 for everyone in about four hours. Levels 2 and 3 earned by producing real work, scored against a rubric. No certificate for watching videos.
What changed, then hands-on
Seventy-five minutes on Teams or Zoom with a practitioner who uses AI in your kind of role every day. Recorded within 24 hours.
One screen for the board
Level distribution by department, task scores, attendance and hours saved. The renewal conversation, in one view.
Three months, one team, a number to show.
The pilot takes five to 25 people through Level 1, the whole course: the cases, the practical, the badge. There is a live session every month, a short brief every week, and the dashboard from day one. It ends with a debrief on the numbers, and $1,500 of the fee is credited if you continue.
See the pilotWhat you get at day 90
Level 1 certified across the pilot group, two months of hours logged, and attendance at the live sessions, all on the dashboard. The first screen for your board, and the decision on the year.
How it works
Three parts, one subscription. The dashboard is what you buy; the ladder and the live sessions make it move.
The ladder
| Level | Who | How it is earned | Time |
|---|---|---|---|
| Level 1 · Fluent | Every employee | A reference module on the tools themselves, then seven cases from one fictional company, decided screen by screen, with an exercise and a quiz in each: what AI does well and badly, how to work with it, and safe use of company data. A twenty-minute practical on an unseen document earns the badge. | About 4 hours |
| Level 2 · Productive | By role: finance, sales, operations, R&D, support, marketing, HR, legal | Seven real tasks per track on realistic fictional data, submitted and scored against a published rubric within five working days. Pass at 80% on all seven. | 10 to 15 hours over about three months |
| Level 3 · Builder | Volunteers who become internal champions | A ten-week cohort. Each person automates a real process in their own company and presents the measured before-and-after to their leadership. | Ten weeks |
Certifications expire after twelve months. Recertifying is a short refresher and one new task, so a badge always means current skills. Fluentia Certified
The monthly live session
New capabilities, tool changes, one thing to stop doing.
One real task, worked through together, in your own tools.
Applied to your company. Counts toward certification.
Private tier sessions run on your workflows and your anonymised data. Trainers are people who use AI in the role, not presenters. Every session is recorded and on the platform within 24 hours.
One employee's first 90 days
A finance analyst, about two hours a week.
- Week 1Kick-off. Level 1 modules 1 to 4. First live session on the calendar.
- Weeks 2 to 3Level 1 done, badge issued. Picks a task from her own week and does it with AI.
- Week 4First private live session: rebuilding one step of your month-end close, together.
- Weeks 5 to 10Finance modules 1 to 5, each submitted and scored with feedback within five working days.
- Week 12Modules 6 and 7 done. Level 2 Productive · Finance badge. Her hours-saved figure is on the dashboard.
The dashboard
What your leadership sees, updated as people progress.
| Metric | Why it matters |
|---|---|
| Staff at Level 1, 2, 3 by department | The ladder, moving |
| Average task score per module | Where the organisation is strong and weak |
| Live attendance and follow-up completion | Whether people are keeping up, not just certified once |
| Hours saved per week, by department | The number that justifies renewal |
| Automations shipped by Level 3 graduates | Tangible internal wins |
Thirty minutes. One finance task, live.
We show the dashboard on sample data, walk through one Level 2 task, and go through the pilot terms in full. You leave with a proposed kick-off date.
Eight tracks, seven real tasks each.
Every Level 2 module is a task from the job, on realistic fictional data, scored against a rubric. The last module of every track is guardrails: catching fabricated figures, wrong dates and data that should never have been shared.
Catalogue updated September 2026. Every module shows its own last-updated date inside the platform; Level 1 modules 1 to 3 are rewritten whenever a major model release changes what AI does well and badly.
Finance
- Clean and reconcile a 3,700-row bank export against the ledger; produce the adjustments log.
- Driver-based forecast from 36 months of history, three scenarios, assumptions listed.
- Variance commentary for the board across eight cost centres; every variance over 5% explained, no invented causes.
- Vendor and contract review: six supplier contracts into one table; the two riskiest flagged and quantified.
- Month-end close accelerator: classify 40 steps, build the first two automations, estimate hours saved.
- AI-assisted Excel: audit a cash-flow model with twelve planted errors; build a debtor ageing schedule.
- Guardrails: audit five AI-drafted board paragraphs against the source; write the checklist a junior would follow.
Sales
- Account research: one-page briefs on twelve targets, facts verified.
- Outreach: a three-touch sequence personalised from the brief.
- Call prep: extract needs and objections from a real transcript, nothing invented.
- Proposal: executive summary and three options from notes and a price sheet.
- Objections: ten real ones, role-played against the model.
- Pipeline forecast: clean a 200-deal export, flag risk, weight the forecast.
- Handover and guardrails: catch a fabricated case study, a wrong price, a promised feature that doesn't exist.
Operations
- SOPs: three processes from walkthroughs into step-by-step procedures.
- Scheduling: a roster from demand, shifts and leave, trade-offs explained.
- Inventory: 1,200 SKUs; find coming stock-outs and over-stock.
- Root cause: cluster eight incidents, find the two recurring causes.
- Reporting automation: design and build the first steps of a weekly report.
- Vendor comms: twenty emails including two disputes; replies and an escalation.
- Improvement and guardrails: catch a wrong unit conversion and a silently dropped safety step.
R&D
- Prior art: a structured review of 25 papers and 10 patents, every claim sourced.
- Requirements: a spec from interviews and a teardown, criteria testable.
- Experiment design: variables, controls, sample size, what would falsify it.
- Data analysis: 50,000 readings with a sensor swap to detect.
- Code assistance: fix three bugs, add tests, catch the AI's wrong fix.
- Documentation: a design note from a lab notebook and commit history.
- IP and guardrails: what may go to an external model; catch a licence violation.
Customer support
- Triage: classify and route 150 tickets, catch the AI's misroutes.
- First replies: twenty tickets, grounded in the manual and refund policy.
- Knowledge base: the ten most repeated questions become help articles.
- Escalation: summarise five tangled threads for engineering or finance.
- Churn signals: score sentiment across thirty accounts, flag six at risk.
- QA at scale: build a rubric, apply it, check the AI's scoring.
- Guardrails: catch a made-up feature, an off-policy refund, another customer's data.
Marketing
- Audience insight: personas and objections from interviews and a survey.
- Messaging: value proposition, pillars and proof points per persona.
- Content at volume: blog, landing page, nurture and social, on brand voice.
- Campaign analysis: two years of channel data; what drove pipeline.
- SEO: cluster 2,000 keywords by intent, write five briefs.
- Creative briefs: generate variations, choose three and defend them.
- Brand and legal guardrails: catch a fabricated quote and an unprovable claim.
HR and people
- Job descriptions: three roles from a manager's email, with scorecards.
- Fair screening: summarise 60 applications; find the bias in the AI's ranking.
- Onboarding: a 30-day plan from the handbook and team walkthroughs.
- Policy and FAQ: a hybrid-work policy and answers to 80 real questions.
- Engagement analysis: theme 300 survey responses by department.
- Workforce data: where attrition concentrates and what it costs.
- Guardrails: catch a discriminatory criterion and personal data pasted externally.
Legal
- Contract review: six agreements against the playbook, with redlines.
- Drafting: from template and deal memo, every departure flagged.
- Research memo: every authority verified; one unverified citation fails.
- Due diligence: a 40-document data room into an issues list.
- Obligations: 80 agreements into a tracker; the eight that need action.
- Regulatory monitoring: twenty updates, three jurisdictions, a board note.
- Privilege and guardrails: what may go to an external tool; catch a fabricated case.
Level 1 for everyone
Nine modules, about four hours, then a twenty-minute practical: what these tools are, what they do well and badly, how to work with them, your company's data policy and approved tools, how to check output, your own daily workflows, and how to keep up. Module 5 is built from your own AI policy. Managers add a 90-minute "Leading with AI" module.
Level 3 for the champions
A ten-week cohort for people who finished a Level 2 track. Pick a recurring process, design the automation, build it with a weekly clinic, harden it with a runbook, present the before-and-after to your leadership. It passes only if it runs on real work and a colleague can operate it.
Thirty minutes. One finance task, live.
We show the dashboard on sample data, walk through one Level 2 task, and go through the pilot terms in full. You leave with a proposed kick-off date.
The pilot, then two tiers.
Start with three months and one team. Continue per seat, per year, with $1,500 of the pilot fee credited.
Pilot
$4,500Three months, flat fee.
- Five to 25 people: Level 1 end to end
- A live session every month, shared with the other pilot companies
- A short brief every week
- The dashboard from day one, and a written reading of it at day thirty and day sixty
- A debrief at day ninety, with a written reading of the numbers
- $1,500 credited against a one-year subscription
Book a pilot call
$1,500 credited if you sign for a year within thirty days of the debrief. The debrief happens either way.
Core
$240Per seat, per year. Minimum 25 seats.
- Platform and all tracks
- Shared monthly live sessions
- Leadership dashboard
- Weekly brief
- Annual recertification
Private
$480Per seat, per year. Minimum 50 seats.
- Everything in Core
- Twelve company-specific live sessions a year
- Quarterly capability review with leadership
- Your own workflows and anonymised data in sessions
| Add-on or term | Detail |
|---|---|
| Level 3 Builder cohort | $2,500 per person, ten weeks, twice a year |
| Volume | 10% off from 250 seats, 20% from 500 |
| Two-year commitment | 15% off, paid annually |
| Extra private session | $1,500 |
| Invoicing | Annual, in advance; CAD, EUR or MXN at the rate on the invoice date |
A 300-person company on Private pays $144,000 a year: twelve practitioner sessions, a certified workforce and a dashboard, for about the cost of one analyst. For context, a half-day workshop for 25 people runs $3,000 to $12,000 and practitioner cohort programmes $2,500 to $6,000 per person.
Questions we usually get
We already have a learning library. Why this?
Libraries teach generically and few people finish them. Fluentia certifies people on their role's real tasks, keeps them current every month, and gives you a number.
Which AI tools does it cover?
The ones you have approved. The skills transfer across tools; the monthly session covers what changed in whichever ones you use.
How much time does it take?
Level 1 is about four hours, plus a twenty-minute practical. Level 2 is ten to fifteen hours over about three months. Live sessions are 75 minutes a month.
Who scores the tasks?
A practitioner in the role, against a published rubric, within five working days. AI helps with a first pass; a person signs every certification.
What if the tools change?
That is the point. The weekly brief and the monthly session exist for it. Modules are reviewed quarterly and show their last-updated date.
Does this cover the EU AI Act?
Article 4 asks for documented, role-calibrated AI-literacy effort, enforceable since August 2026. The ladder and dashboard are that record. It is a by-product of the programme, not its purpose.
What about managers and leadership?
A 90-minute "Leading with AI" module for anyone who manages people: what to expect from a fluent team, how to read the dashboard, how to review AI-assisted work. Every pilot opens with a kick-off for its sponsor.
Where do our people's questions go?
Three places. A FAQ open to everyone taking the course, the weekly brief, which answers one of the questions received each week, and the live sessions. A question is published without the person's name or the company's.
Does the certification mean anything outside the company?
It is verifiable online and carries the module scores. It is new; its weight grows with the companies that use it. Inside the company it means this person's work can be trusted with AI.
Thirty minutes. One finance task, live.
We show the dashboard on sample data, walk through one Level 2 task, and go through the pilot terms in full. You leave with a proposed kick-off date.
Your data, your tools, your policy.
What IT and legal will want to know before anyone signs.
We teach your rules, not ours
Level 1 module 5 is built from your AI use policy and approved tool list. If you don't have one yet, we provide a template and help you adopt it in week 1.
Training datasets are fictional
Every Level 2 task uses realistic but invented company data. Nobody practises on your ledger.
Private sessions use anonymised data
When we work on your real workflows, you choose what is shared and it is anonymised first. Recordings are stored in your workspace only.
For customers in France
Article 4 of the EU AI Act has been enforceable since 2 August 2026. It asks for documented, role-calibrated AI-literacy effort, not a blanket course. A role-based ladder with a dated dashboard is that record. We don't sell compliance; you get the documentation as a by-product.
Platform data
Sign-in with a work email, verified by a link. Progress and scores belong to you and export on request. Hosted in Canada. Graded answers go to a grading model to be scored: Fluentia does not keep them, and the provider does not use them to train its models. Data processing agreement available; Quebec Law 25 and GDPR aligned.
What a manager sees
By name: progress, badge, live sessions attended and last activity. Scores and hours appear by department only, for five people or more. Every learner can read exactly that on their own account page.
Who is behind Fluentia
Fluentia is run by its founder, a practitioner with a finance and technology background, who leads the live sessions personally during pilots.
Trainers are practitioners who use AI in the role every day: a controller for finance, a sales lead for sales, an operations manager for ops. Content is reviewed every quarter and rewritten when the tools change.
Verify a certificate
Every Fluentia badge carries a verification code. Enter it to see the name, company, level, issue date and expiry.
Codes are checked on the platform; pilot-issued codes can also be confirmed by email at [email protected].
Thirty minutes. One finance task, live.
We show the dashboard on sample data, walk through one Level 2 task, and go through the pilot terms in full. You leave with a proposed kick-off date.
Book a pilot call.
Thirty minutes. We show the dashboard, walk through one finance task live, and go through the pilot terms. You leave with a proposed kick-off date.
On the call
The dashboard on sample data. One finance module, live. The pilot terms, in full. Your questions. A proposed kick-off date.
Pick a time
Thirty minutes, Eastern time (Toronto/Montréal).
Privacy
Fluentia collects only what the pilot call form asks for, uses it to prepare and follow up that call, and deletes it on request. No advertising cookies, no third-party trackers. Analytics, if any, are cookie-free and aggregated. Data processing agreement available for customers. Contact: [email protected].