AI consultancy for mid-market and enterprise companies, from strategy to implementation.

Strategy, implementation, governance and enablement from one senior team, so the plan and the build never sit with different firms.

A boutique, engineering-led AI consultancy.

Twistag helps mid-market and enterprise companies find where AI is worth the investment, then builds the systems and stays with the team until they are in daily use. Consultants and engineers work as one team of 25, from Lisbon, since 2016.

How an engagement runs
  • DHL
  • Volkswagen
  • Autodesk
  • Bangkok Bank
  • Ruinart
  • Glovo
  • Nike
  • Worten
  • Defined.ai
  • Indie Campers
What it covers

What AI consulting at Twistag covers.

Four parts of one engagement. A company can start at any of them, and most start with the diagnosis.

  1. 01

    The
    diagnosis.

    The AI Readiness Sprint: operations, data, systems, people and governance, reviewed in 2–4 weeks from €5,000.

  2. 02

    Roadmap and
    business case.

    The use cases worth funding first, each with a business case, its risks and a 90-day plan leadership can approve.

  3. 03

    The
    build.

    AI agents, data platforms and AI-native products, built inside the company's own systems and tested against an agreed number.

  4. 04

    Operate and
    enable.

    AI Workforce Enablement for the wider team, and forward deployed engineers who run and improve what was built.

How it runs

How an engagement runs.

Some companies arrive with a use case, others with a question. Both start with the same sprint, and commitment grows only after the first system proves its number.

Starts with the AI Readiness Sprint.

Two to four weeks on how the work runs today, from €5,000. The sprint ends with a shortlist of use cases ranked by payback, a business case and a 90-day plan leadership can approve.

First system in 6–8 weeks.

One use case, shaped around the systems, approvals and habits already in place, in daily use and measured against the number agreed in the sprint.

Scale on proof.

The next use cases follow once the first one holds its number. Commitment grows one step at a time, and each system builds on the last.

Handover or forward deployed engineers.

The internal team takes the system over with the code, documentation and training, or forward deployed engineers stay to run and improve it.

By role

What each seat at the table gets.

01 · For CEOs

Where AI changes the economics, and in what order.

A roadmap ranked by payback and tied to revenue, margin and headcount plans, with one team accountable from the plan to the running system.

A ranked roadmapUse cases ordered by payback, risk and effort.
A board-ready caseEach use case with its cost, return and risks.
One accountable teamThe people who plan the work build and run it.
Commitment on proofEach next step funded by the last one’s number.
AI strategy consulting
02 · For CFOs

What each investment returns, and when.

A fixed fee to start, a business case per use case, and every system measured against a number agreed before the build.

A fixed-fee startThe AI Readiness Sprint, from €5,000.
A case per use caseCost, return and payback period, before any build.
Phased commitmentBudget follows results, one step at a time.
Measured resultsEach system tracked against its agreed number.
AI Readiness Sprint
03 · For CIOs and CTOs

How AI fits the stack, the security model and the data.

Systems built in the company's own cloud and repositories, under the same controls as the rest of the estate.

Own cloud and codeRepositories, infrastructure and models owned from day one.
Security and complianceAn ISO/IEC 27001 practice and EU AI Act controls.
Traceable agentsEvery agent decision can be explained, audited and reversed.
Model choice per taskEach step uses the model that passes its tests at the lowest cost.
AI governance and EU AI Act readiness
04 · For COOs

Which workflows AI can carry, and what stays with people.

Agents take the reading, checking, matching and chasing. People keep the exceptions and the decisions.

A workflow inventoryWhere the hours go today, task by task.
Agents in the flowBuilt into the systems and approvals already in place.
People in controlExceptions and approvals stay with the team.
Adoption that holdsTraining and usage tracked until it sticks.
AI automation consulting
Compared

What engineering-led changes.

Three ways to buy AI consulting compared: Twistag, a strategy-led consultancy and a build-only agency.
Compared onTwistagStrategy-led consultancyBuild-only agency
Who writes the planConsultants and engineers, togetherStrategy consultantsThe client
Who builds itThe same teamA separate delivery team or partnerThe agency, to the client's brief
The first weeks produceA ranked roadmap, a business case and a 90-day planA strategy and a recommendationA prototype
First system live6–8 weeks after the sprintAfter a separate build phaseSet by the brief
Measured againstA number agreed before the buildProgramme milestonesDelivery to spec
Who owns the codeThe client, from day oneAgreed per contractAgreed per contract
In their words

What clients say after the work ships.

We have relied on Twistag as a partner for quite a few years. They are a really good partner with a great team. They always deliver on expectations and are very responsive to our needs when they change. It is really hard to deliver quality year after year but Twistag does exactly that.
Stefan LavénFounder & CEO, Datatalks
Twistag dedication to understanding our business goals and their innovative approach to solving challenges were impressive. They were highly responsive to our needs and operated as a true partner.
João SilvaBusiness Manager, Glovo
Twistag team are really easy to work with, they helped us build our first subscription-based e-commerce platform which helped us test our product with our early users and get market feedback fast.
Tomas FroesFounder & CEO, kencko
Twistag has delivered a ready-to-use MVP on time that received positive great feedback. The team has exceptional project management skills, excellent communication, insightful feedback, and dedication. Their proactiveness and profound understanding of the client’s product and business are impressive.
Ana RitaChief Scientific Officer, VOH.CoLAB
Our MVP was developed for a small target audience, validated, and refined into a subscription-based product and logistics solution. In 12 weeks, Twistag delivered the company website, mobile app, and inventory management integrations, along with a seamless onboarding and training process.
Ricardo AlvarezCo-founder & CEO, Keepwhat
Related work

From the roadmap to the running system.

All case studies
  • DeloitteTechnology Fast 500 EMEA2023 and 2024
  • Tech BehemothsWinner, Artificial IntelligencePortugal, 2025. Custom Software Development, Portugal, 2024
  • Clutch4.9 out of 5From 14 verified client reviews
  • AWSPartnerCloud, data and AI services
  • ISO/IEC 27001Information securityPractice behind every engagement
In the press
Ways to work

Three ways to work together.

For mid-market and enterprise companies in Europe, North America and Asia, and the PE operating partners who roll AI out across a portfolio. Most start with the AI Readiness Sprint.

  • Embedded

    Forward Deployed Engineering.

    Senior engineers inside the client's team, in its repo, CI and backlog, priced per engineer-day and resized each quarter.

    See how a pod works
  • Fixed scope

    Fixed-scope project.

    One system, delivered to spec: an agreed outcome, price and date.

    See the ways to work
  • With partners

    Partner delivery.

    Tech delivery for consulting firms’ clients, senior capacity for a tech company’s own roadmap, and AI playbooks rolled out across PE portfolios.

    Explore a partnership
Services

AI consulting services.

Questions

About AI consulting at Twistag.

What does AI consulting cost?

It starts with the AI Readiness Sprint: a fixed fee from €5,000 for two to four weeks. The build that follows is priced on its scope, as a fixed-scope project or per engineer-day for forward deployed engineers, and agreed before work starts. A written proposal follows within a week of a 30-minute call.

How is Twistag different from a large consulting firm?

Twistag is a boutique AI consultancy that takes mid-market and enterprise companies from AI strategy to implementation, with forward deployed engineers working inside the client's team. The senior team of 25 is small enough that the people in the first meeting are the people who build, so no handover sits between the roadmap and the system. It is tool-agnostic, and each engagement grows only after the first system proves its number.

Does Twistag build or only advise?

Both, in one team. The sprint ends with a roadmap and a business case; the same people then build the first system and run it until the internal team takes it over. The roadmap can also be delivered by an internal team or another partner.

How long until the first system is live?

Six to eight weeks after the AI Readiness Sprint, which itself takes two to four weeks. The first system covers one use case and is measured against the number agreed in the sprint.

Which industries?

Financial services, retail and e-commerce, hospitality, healthcare, manufacturing, logistics, utilities and media. Clients include DHL and Volkswagen. The industry pages set out the work sector by sector.

Can Twistag work alongside an internal team or another consultancy?

Yes. Forward deployed engineers work inside the client's team, in its repo, CI and backlog. For consulting firms, Twistag delivers the technical work for their clients through partner delivery, under the consultancy's lead.

Where is Twistag based and which markets does it serve?

Lisbon, Portugal, since 2016. Twistag works with clients across Europe, North America and Asia, remotely and on site when the work needs it. Lisbon shares the full European working day and the morning of the US East Coast.
Talk to us

Start with a scoping call.

Thirty minutes with a senior lead on the company, the work that should change first and what a sprint would cover.

A senior lead replies within one working day.

AI consulting

Tell us where work gets stuck.

A senior engineer replies within one working day.

Interested in

Thirty-minute call

Pick a time.