Personal AI consulting · Teams of 10 to 10,000
Lift every employee. Prove every lift.
Upskill your employees and scale your business with individual, personalized AI consulting. Every person gets their own analysis, their own AI setup and hands-on help building their own tools. Then we measure the result against their own baseline.
- One person at a timeEvery plan starts from one person's real week.
- Measured per personEach lift is timed against that person's own baseline.
- Vendor-neutralCopilot, ChatGPT, Claude, Gemini or open models. We set up what fits.
- Built to hand overYour people learn to build their own tools, so they can keep improving after we leave.
The problem
Most Danish companies use AI now. Few can show the lift.
Use is climbing fast. Far fewer companies can show what it changed for each person, or what it was worth. The technology works on the right tasks. In our view, generic rollouts miss them because they ignore how each person actually works.
59%
of Danish companies with 10 or more employees used AI in 2026, up from 15% in 2023.
Statistics Denmark, ITAV28 (2026), private sector excl. finance
28%
of managers and executives at large US companies report strongly positive returns from AI, although 80% report some return.
Wharton & GBK Collective, The Adaptive Advantage (October 2026)
36%
of workers surveyed feel properly trained in AI, unchanged since 2025, while 88% expect to need major upskilling within five years.
40%+
higher-quality work from consultants using AI on tasks it suits. On a task it didn’t suit, they were 19 percentage points less likely to get it right.
Dell’Acqua et al., HBS & BCG, 758 consultants, GPT-4, 2023 (Organization Science, 2026)
Fit is the missing piece. Each person needs the right setup for their own tasks, and a measurement that proves it worked.
The difference
Built around each person's actual week.
In our view, most programs teach a tool to a room. We start from one person's work, set AI up around it, and measure what changed.
| Aspect | Typical rollout, as we see it | LØFT |
|---|---|---|
| Built for | Everyone at once | One person at a time |
| Starts from | The tool's feature list | Each person's week, timed |
| People leave with | Slides and a certificate | A working setup and tools they built |
| Proof of success | A satisfaction survey | Measured lift against their own baseline |
| Technology | One default setup | Tuned per workflow: fast, frugal or deep-thinking |
| After the consultants leave | Momentum fades | Champions and a growing tool library |
The method
Four phases, from the boardroom to every desk.
We read the whole business first, go deep on every person, build the foundation with IT and management, and then coach each employee until your organization improves on its own.
- Phase 01Map2–5 weeks
Strategic surface analysis of the business
What happens
- Leadership and department interviews
- Processes, systems and data mapped
- AI readiness, security and compliance check
- Gap analysis and an opportunity map ranked by value
You get
- Opportunity map
- Scoped plan
- Agreed lift target and method
Who's involved: leadership, department heads and IT
- Phase 02Measure2–6 weeks
A personal deep-dive for every employee
What happens
- A guided 20-minute AI intake interview for each employee
- Key workflows timed, from system data or short time samples, to set personal baselines
- Roles grouped into archetypes, then tailored person by person
- A personal Lift Plan with each person's estimated upside
You get
- Personal Lift Plans
- Baseline per person
- Upside forecast per team
Who's involved: every participating employee
- Phase 03Build3–8 weeks
Foundation first, then every person
What happens
- With IT and management: platform, security, permissions, data connections, usage policy and cost controls
- The AI harness tuned per use case for speed, cost or deep reasoning
- A shared library of skills and prompts for your company
- A personal setup for each employee, and 1:1 build sessions where they build their first tools
You get
- Production-ready AI environment
- Governance and policy
- Personal setups
- First tools in daily use
Who's involved: IT admins and management first, then every employee
- Phase 04CompoundMonthly · optional
Ongoing upskilling and advisory
What happens
- Individual coaching cycles and open office hours
- A trained champion network that coaches peers
- New use cases and builds every quarter
- Lift re-measured and reported, plus leadership advisory as models and tools change
You get
- Monthly lift report
- Growing tool library
- An organization that keeps improving
Who's involved: champions, teams and leadership
The promise
Our name is a promise.
In Danish, et løft is a lift and et løfte is a promise. We make one about the other.
- Step 1TargetWe agree a target of at least 20% time back on each participant's targeted workflows.
- Step 2BaselineBefore we change anything, we time that work.
- Step 3MeasureAbout 60 days after each person's setup, we time the same workflows again.
- Step 4Fee at riskPart of our fee is invoiced only when the target is met.
Lift means time saved, or more output in the same time, on in-scope workflows, compared with each participant's own baseline. Exact terms are agreed per engagement. Each person sees their own numbers, and management sees results only for groups of five or more.
The math
What a lift is worth.
Set your own numbers. The default lift is our minimum target.
Capacity gained
160
full-time equivalents of capacity, without a single hire
Each dot is 5 people. Yellow dots show the capacity gained. Illustrative.
Value per year
€15.8M
Hours back per year
264,000
Per person
2.9 h/week
Lift on total capacity
8%
Illustrative. Assumes 1,650 working hours per person per year and 46 working weeks. In the Measure phase we replace the time assumptions with your measured baselines.
Lift Plans
Every employee gets a personal Lift Plan.
This is what one looks like. Pick a role.
Sales
Account manager
B2B, 6–10 customer meetings a week
Example · illustrative
7.5 h
back every week
58%
less time on these workflows
20%
of a 37-hour week
After setupTime backHours per week, scale 0–10
Their setup
- Meeting assistant connected to calendar and CRM
- Research assistant that writes a one-page account brief
- Proposal skill built from your won deals and price list
They learn to build
An account-brief agent that pulls CRM history, recent news and open support tickets before every meeting.
Scale
Personal at 10 people. Personal at 10,000.
The method stays the same at every size. Only the way we reach everyone changes.
10–50 people
Team
Every person gets 1:1 deep-dives and 1:1 build sessions.
Plan for 6–8 weeks
50–500 people
Company
1:1 deep-dives for everyone, build clinics by role, and trained champions in every team.
Plan for 8–14 weeks
500+ people
Enterprise
A measured pilot cohort first, then rollout in waves, with an AI intake for every employee and a champion network.
Plan for 3–9 months
How it stays personal at 2,000 people
We use the same AI we teach. It lets a small senior team give every employee individual attention.
- 1An AI intake for everyoneA guided 20-minute interview maps each person's week, in Danish, English or German.
- 2From archetypes to individualsWe group similar roles, then tailor each plan to the person.
- 3Hands-on time for everyoneEvery person builds with a coach, 1:1 or in a small clinic with others in the same role.
- 4Champions multiply itAbout one trained champion per 25 employees, to spread what works.
- 5Measured wavesEach wave is measured before the next one starts.
Start with one team.
One team. We inform employees first, take baselines, set everyone up and measure again about 60 days later. Same promise, price agreed up front.
From the lab
We build. Then we teach your people to build.
Examples of what we can build with your people, or for them when it's too niche to learn.
Meeting-to-CRM autopilot
Transcribes the meeting, pulls out decisions and next steps, updates the CRM and drafts the follow-up email.
Handbook assistant
Answers policy questions with sources, in Danish and English, inside your own environment.
Proposal generator
Turns a call summary and your price list into an on-brand proposal draft.
Month-end anomaly check
Flags unusual postings and drafts the variance comments before the review meeting.
Engineering harness
Model routing, repository context, token budgets and review rules, tuned per team.
Hardware and robotics prototypes
From idea to working prototype: sensors, 3D-printed parts and AI-written firmware.
The team
Strategy, systems and speed in one team.
Four founders, and every engagement gets all of us.
1 founder
Business strategist
Reads the whole business before anyone touches a tool: strategy, P&L, processes and people. Makes sure every lift lands where it moves the numbers.
Leads Map · leadership advisory
2 founders
AI systems architects
Senior technologists who set up and tune the full AI harness: models, routing, context, connectors, permissions and cost. Fast where speed matters, frugal where volume matters, deep-thinking where quality matters.
Leads Build · IT and security
1 founder
Venture and product lead
Gets to the core of a problem fast and builds working versions quickly: skills, automations, full systems and hardware prototypes. Then teaches people to do the same.
Leads Measure and Compound · builds and coaching
Questions
What leaders ask us first.
How is this different from AI training?
Training is one ingredient. Every person also gets their own analysis, their own setup and live build time, and we measure what changed for them.
We already have Copilot, ChatGPT or Gemini. Do we need new tools?
Not necessarily. We're vendor-neutral and start from what you have. We tune it, connect it to your data, and add tools only where they pay off.
How do you measure the lift?
We time each person's targeted workflows before setup and again about 60 days after. Where your systems already log the work, such as tickets, CRM activity or code, we use that data, once employees have been told and only where that data may be reused. Elsewhere we use short, structured time samples. We don't rely on how people feel about it. In a 2025 trial with early-2025 tools, experienced developers took 19% longer with AI while believing it made them 20% faster. METR, which ran the trial, now marks that result as out of date, and self-reports remain a weak measure.
Is per-person measurement a form of surveillance?
It's there to improve each person's setup, and it's never used to rate people. Each person sees their own numbers, and management sees results only for groups of five or more. We agree the method with management, IT and employee representatives up front and inform employees before any measurement starts. We don't use keystroke logging, screen capture, webcams or emotion analysis. This is general information, not legal advice.
Is our data safe?
IT and security come first, before anyone gets a setup. We work inside your environment and policies, use enterprise configurations where your data isn't used to train models, and give every connection its own permissions.
Does this help with the EU AI Act?
Since February 2, 2025, Article 4 of the AI Act has required organizations that provide or use AI systems to take measures on their staff's AI literacy. The Digital Omnibus on AI, in force since July 27, 2026, reworded the duty. Organizations must now take measures that support their staff's AI literacy, and no specific level has to be guaranteed. National authorities supervise it from August 2, 2026. A documented, role-based program like ours is one such measure, and you keep the records. This is general information, not legal advice.
What happens when you leave?
That's the plan. Your champions, your tool library and your people's skills stay with you. Ongoing advisory is optional.
What does it cost?
Pricing depends on scope and headcount, with part of the fee tied to the measured lift. A 30-minute scoping call gets you a number.
Get started
Start with one team. Prove it. Then scale.
Tell us how many people, which roles, and where the time goes today. We'll come back with a pilot proposal and a measurable target.
Email us
hello@loeftconsult.com
Useful to include
- Team size and the roles in scope
- The AI tools you already pay for
- Where the hours go today