AI integration & product studio · Barcelona

Put AI to work in your business.

We help companies automate repetitive processes, connect AI to the tools they already use and ship AI products their teams rely on. Senior engineers, measurable gains, zero hand-offs.

Forward-deployed AI engineering

Less busywork. More output.

We work like forward-deployed engineers: embedded with your team, inside your systems and data, from the first workflow worth automating to AI running in production. Then we stay to measure and improve it, keeping a person in control where it counts.

  1. 01

    AI agents & automation

    Agents that read, decide and act across your tools (CRM, ERP, inbox, spreadsheets) to finish multi-step work like reconciliations, triage and reporting, with a person approving what matters.

  2. 02

    Copilots for your team

    Assistants that prepare drafts, answers and decisions for people to review and approve, so every expert gets through more in less time.

  3. 03

    Knowledge assistants (RAG)

    Ask your documents, tickets, policies and data in plain language, with answers grounded in your own sources and citations, built on retrieval pipelines and vector search.

  4. 04

    Integrations & MCP

    Connect models securely to your internal systems and APIs through tool calling and MCP servers, with the permissions and audit trails your IT team expects.

  5. 05

    AI in your products

    LLM features inside the apps and platforms you already run: matching, summaries, recommendations, coaching, voice and chat.

  6. 06

    Document & data extraction

    Turn invoices, CVs, contracts, forms and emails into structured data your systems can use, with no manual retyping.

  7. 07

    Evals, guardrails & monitoring

    Test sets that measure quality before every change, guardrails against bad outputs, and tracking of accuracy, latency and cost per task in production.

  8. 08

    AI strategy & enablement

    Pick the use cases with the clearest return, choose models and vendors without lock-in, and train your team to own and extend what we ship.

How we work

  1. Step 1

    Embed & discover

    We sit with the people doing the work and map the workflows, data and systems behind it.

  2. Step 2

    Prototype on real data

    A working version on your own data early, so decisions rest on evidence rather than slides.

  3. Step 3

    Ship to production

    Integrations, access controls, approval steps and a rollout your team is ready for.

  4. Step 4

    Measure & improve

    Evals, monitoring and cost tracking, then iterate as models and your needs change.

  • A person approves what matters
  • Privacy and consent by design
  • Cost limits on every AI call
  • Any model: OpenAI, Anthropic, Gemini or open source
  • Built on the systems you already use

AI products we’ve shipped

AI already at work.

Copilots, coaches and assistants running in production for finance teams, recruiters, families and fitness coaches.

Selected clientsAcross marketplaces, mobility, food and delivery

How we help

From first signal to shipped product.

Build a product, bring AI into your business or get expert guidance by the hour. Choose the engagement that fits what your team needs next.

01Discovery first

AI agents & integrations

Forward-deployed AI engineering. We work inside your team, your systems and your data to find the workflows worth automating, then ship AI to production: agents that act in your tools, copilots, knowledge assistants over your own documents and integrations with the software you already run, with evals, guardrails and monitoring built in.

05Flexible hours

Expert consultancy

Get direct access to senior expertise when you need it. Work through product decisions, software architecture, code reviews or AI adoption with practical guidance tailored to your team.

Built for momentum

Senior talent, close to the product.

We partner with ambitious teams to build practical AI tools and mobile products that make a measurable difference to their business.

No relay race between strategy, design and engineering. The people shaping the product are the people building it.

Forward-deployed AIAI agents & automationRAG & knowledge basesProduct strategyUX/UI designiOS & AndroidWeb platformsTechnical consultancy

The 8-week MVP

Enough structure to move fast. Enough room to get it right.

Designed for one primary platform, one decisive user journey and a team ready to make daily product decisions.

Check project fit
  1. 01
    Weeks 01–02

    Daily discovery

    Daily working sessions to align on the problem, users, priorities and the smallest coherent release. We leave with a signed-off scope and delivery map.

  2. 02
    Weeks 03–06

    Implementation

    Four focused weeks of product design and engineering, with visible progress and decisions made close to the work.

  3. 03
    Weeks 07–08

    Review & polish

    Two weeks reserved for client review, iteration, quality assurance and the final layer of fit and finish before handover.

More selected work

Products with somewhere to go.

Health and family apps, delivery, marketplaces and more: 16 products for clients from startups to household names.

Start somewhere real

Let’s make your next release matter.

Send a project brief

Usually answered within 2 business days

Your brief goes straight to our team. We’ll review it and follow up with a clear next step.