We make Foundry work for real in your company.

We were Foundry customers for five years. We know the problem because we lived it and the solution because we built it.

Ontology, automation, agents and decision support on Palantir Foundry and AIP.

Who we help

Do any of these sound familiar?

We work with companies at different stages of their journey with Foundry and AI. This is what they usually ask us.

01

You’re just starting with Foundry

  • Where do I start so I don’t waste money?
  • Which use cases will pay off first?
  • Do I need Foundry for everything, or can some things be done outside it?

What we do for you: a process map, so the shine of the tool doesn’t blind you and you start with what really matters.

Process map

02

I have a clear use case and I want it working

No more glorified demos: a project with fixed scope, timeline and price.

  • How much does it cost to automate my finance team, or my business intelligence?
  • Who will build it from start to finish, not in loose hours?
  • How do I know it will be used and will move my numbers?

What we do for you: we assess a specific use case with you, give you a fixed price and build it from start to finish until it runs in your operation, with its adoption and impact measured.

Fixed-scope project

03

Your AI pilot worked, but now what?

Let the bots do the bots’ work, and people only the work that needs a person.

  • How do I go from the demo to real operations?
  • How do I make agents really save work?
  • How do I get the team to use it every day?

What we do for you: a fixed-scope project and agent deployment on your real workflows, until the savings are measured in hours and cost.

Agent deployment

04

I want to train my team

So the knowledge stays in-house and we depend less on outsiders.

  • How do I get my team to build in Foundry on their own?
  • How do I stop the expertise leaving when the project ends?
  • Who supports my team while they learn?

What we do for you: training on your own data and support until your team can do it alone. The expertise stays in your company, which is where it belongs.

Training and knowledge transfer

05

I want to cut my Foundry costs

The platform bill is growing faster than the value it gives us.

  • Why are my ontology and AIP calls going through the roof?
  • How do I reduce consumption without losing capabilities?
  • Which simple decisions could a model of our own handle?

What we do for you: we audit consumption and cut it: fewer ontology calls, fewer search rounds, AIP only where it adds value and local models for simple decisions.

Cost optimisation

06

I know what I want to do, but I’m short of people

My roadmap is clear and there’s more work than my team can get through.

  • How do I clear the backlog without opening five hiring processes?
  • Where do I find Foundry people who deliver from day one?
  • Who will roll up their sleeves and build with my team, not just give opinions?

What we do for you: an hours bank of senior FDEs who join your team and get stuck in with you: they build, deploy and maintain at the pace you set.

Hours bank

What we do

Six services. You can start with any of them.

Map, project, deployment, training, cost optimisation or hours bank. They connect, but each one works on its own.

01

Process map

Mapped processes, an assessment of your stack, an initial ontology and a prioritised roadmap. It’s yours even if we don’t continue.

See the service

02

Fixed-scope project

One use case from start to finish, with adoption and impact measured, and scope, timeline and price fixed up front. Includes SaaS Consolidation.

See the service

03

Agent deployment

On top of a workflow that already works, agents that do the repetitive work and leave people the work that needs judgement.

See the service

04

Training and knowledge transfer

Your team learns Foundry and AIP working with their own data, at the end of the project or continuously.

See the service

05

Cost optimisation

Fewer ontology calls, fewer search rounds and AIP only where it adds value. Local models for simple decisions.

See the service

06

Hours bank

To maintain and run what we have built, or as free hours for whatever you need.

See the service

Not sure where to start? We’ll work it out on a call.

Book a Discovery Call

How we work

We start with the decision, not the tool.

First we understand what needs deciding and who decides it. The technology comes after.

01

Use case

Which business problem we solve and what it’s worth.

  • Finance
  • Operations
  • Business intelligence
  • Human resources
  • Cybersecurity
  • Communications
  • Procurement
  • Marketing

02

Decision

Which specific decision sits behind it, and who makes it.

  • 1C-suiteWhere to invest, expand or cut.
  • 2Middle managerWhat to prioritise, who to assign it to and when to review it.
  • 3OperatorWhat to do now, with the information already digested.

03

Human or machine

Which part a person decides and which part an agent can do.

Commercial models

  • AIP
  • Claude
  • ChatGPT

Open-weight models / Sovereign AI

  • Hugging FaceGemma · Qwen · GLM

Bots do the bots’ work. People do the work that needs judgement.

04

Ontology

The objects and actions that decision needs.

Palantir Foundry

  • Objects
  • Links
  • Actions

05

Data and systems

We process and model the data, and connect to the systems you already have.

  • Excel
  • Word
  • Postgres
  • Elastic
  • Google Drive
  • Python files
  • SAP
  • Salesforce
  • HubSpot
  • Outlook
  • Teams
  • Slack
  • Jira
  • Notion
  • In-house legacy

About us

years of Foundry experience
per FDE

  • Senior profiles only
  • You work directly with the people who build

We know the problem because we lived it. We know the solution because we built it.

We have been on the other side: years of deployment without knowing what came next, bets that failed and use cases that worked. We don’t sell smoke. We get our boots muddy.

Each of our FDEs has spent more than five years designing ontologies, integrating complex systems and putting solutions into production that are now part of the daily operations of demanding organisations.

Our ambition: to bring together the most experienced team of Foundry-specialised FDEs on the market.

We start at home

We are modelling our own operation in Foundry.

Before modelling a client’s operation, we model our own: every Astrolabe process, from the first call to the invoice, connected in an ontology on Foundry.

Noticia Empresa Oportunidad Propuesta Correo Persona Proceso Proyecto Reunión Decisión Cliente Factura

Demos

Applications in Foundry

Demos of applications we have built are coming soon.

Want to know what we could build for your company?

Book a Discovery Call

Articles

What we have learned, in writing.

Honest comparisons, mistakes we have all made and anonymised cases. Coming soon.

  • Tools

    Do you need AIP for this? An honest comparison with n8n and other alternatives

    Read the article
  • Ontology

    The ontology mistakes we have all made in Foundry

    Read the article
  • Operations

    How to design alert workflows people actually use

    Coming soon
  • Geo

    All the Foundry Geo features you didn’t know existed

    Coming soon
  • Sovereign AI

    Make your AI sovereign, and cheaper too

    Coming soon
  • Evals

    You have a gold mine in your logs and you’re giving it away to OpenAI

    Read the article
  • Adoption

    Adopting AI isn’t giving everyone Copilot

    Read the article
  • Decision

    Forget the Data Scientist: what your company needs is a Decision Scientist

    Coming soon

Questions

What people usually ask us.

  • Do I need Palantir Foundry?

    We mostly work on Foundry. If the company doesn’t have it yet, the mapping checks whether it is the right tool and, if not, we look for alternatives.

  • How long does the process map take?

    Between 2 and 4 weeks, depending on the number of areas and people involved.

  • How do you charge?

    By project, with scope, timeline and price fixed up front, or with an hours bank: maintenance hours for what we have built, or free hours for whatever you need.

  • What happens when a project ends?

    Your internal team can carry on alone, train with us, or rely on an hours bank for changes and support, with agreed response times.

  • Do you replace my internal team?

    No. Your internal team works with us from day one so they can run the system on their own.

  • Which systems do you connect to?

    Whatever already exists: ERPs such as SAP, databases, spreadsheets, CRMs, email and documents.

  • Where do you work?

    From Spain, for organisations in any country. We work remotely and travel when the project needs it.

Book a Discovery Call

We reply within two working days. The first conversation is for understanding the problem before proposing anything.

We only use this data to reply to you. See the privacy policy.

Request ready.

Your email app has opened with the message addressed to hello@astrolabes.io. Send it and we will reply within two working days.

Choose the slogan

The hero proposals. Your choice is remembered in this browser.