# HeritageTech, together: how our open-source stack could plug into UAlgTec Heritage, the University of the Algarve's new European centre of excellence for heritage science and technology, mapped from the dig to the museum, with encrypted evidence vaults per site, signed finds, graphs of every excavation, agents with written policies, and what I would write in its "start a conversation" form, sgit.ai

> Today I saw a screen with an invitation: "Help build what comes next. Bring a problem. Bring a technology. Bring an idea." It is from UAlgTec Heritage, UTH, the University of the Algarve's new centre of excellence, coordinated through its ICArEHB research centre and led by João Cascalheira. It won about 31.5 million euros, 15 million of it from Horizon Europe's Teaming for Excellence, the first such award in archaeology in Europe, with Kiel University, ERA Arqueologia and Zartis as partners, and it plans to start in January 2027. Its aim is a European HeritageTech sector, like FinTech or HealthTech, built on an AI-first platform for archaeological heritage. That is very close to what we have been building. An AI-first platform for heritage needs what every AI-first platform needs: evidence that cannot be quietly changed, records signed by the person who made them, data from many teams that does not collide, agents that only do what they are allowed to, and claims a reader can trace back to the find. So here is the mapping, stage by stage, from the field to the archive to the public; what UTH brings and what we bring; a first pilot on one site; and the answer I would give in its form.

*Source: <https://newsroom.sgit.ai/articles/heritage-tech-together.html> · sgit.ai v0.7.87 · this file is generated from the same content as the page, so the two cannot drift. Every page on this site has a `.md` twin; internal links below point at them.*

---

[Home](../index.md) / [Articles](index.md) / HeritageTech, together: how our open-source stack could plug into UAlgTec Heritage, the University of the Algarve's new European centre of excellence for heritage science and technology, mapped from the dig to the museum, with encrypted evidence vaults per site, signed finds, graphs of every excavation, agents with written policies, and what I would write in its "start a conversation" form

# HeritageTech, together: how our open-source stack could plug into UAlgTec Heritage, the University of the Algarve's new European centre of excellence for heritage science and technology, mapped from the dig to the museum, with encrypted evidence vaults per site, signed finds, graphs of every excavation, agents with written policies, and what I would write in its "start a conversation" form

By [Dinis Cruz](https://sgit.ai/about/index.md) · 2026-11-17 · [article v1.0.0](versions/heritage-tech-together.md) · [newsroom v0.14.140](https://newsroom.sgit.ai/admin/versions.html) · heritagearchaeologyalgarveresearchprovenancevaultsgraphspartnershipsarticle

***Abstract:** Today I saw a screen with an invitation: "Help build what comes next. Bring a problem. Bring a technology. Bring an idea." It is from UAlgTec Heritage, UTH, the University of the Algarve's new centre of excellence, coordinated through its ICArEHB research centre and led by João Cascalheira. It won about 31.5 million euros, 15 million of it from Horizon Europe's Teaming for Excellence, the first such award in archaeology in Europe, with Kiel University, ERA Arqueologia and Zartis as partners, and it plans to start in January 2027. Its aim is a European HeritageTech sector, like FinTech or HealthTech, built on an AI-first platform for archaeological heritage. That is very close to what we have been building. An AI-first platform for heritage needs what every AI-first platform needs: evidence that cannot be quietly changed, records signed by the person who made them, data from many teams that does not collide, agents that only do what they are allowed to, and claims a reader can trace back to the find. So here is the mapping, stage by stage, from the field to the archive to the public; what UTH brings and what we bring; a first pilot on one site; and the answer I would give in its form.*

From the dig to the public: what an AI-first heritage platform needs at each stage, and the open piece we already have.

## In short

- **UTH is a new European centre of excellence** for heritage science and technology, at the University of the Algarve.
- **Its aim is a HeritageTech sector,** on an AI-first platform for archaeological heritage, from January 2027.
- **An AI-first heritage platform needs proof:** evidence, signatures, provenance, agents with limits.
- **That is what our open-source stack does,** already running in our own newsrooms.
- **Six places it plugs in,** from the field record to the public story.
- **A first pilot:** one site, one season's records, in the open.
- **The first step is theirs to invite and ours to take:** a short answer in their form.

## UAlgTec Heritage

UAlgTec Heritage, UTH, is the University of the Algarve's new centre of excellence, coordinated through ICArEHB, its Interdisciplinary Center for Archaeology and the Evolution of Human Behaviour, with João Cascalheira as director and project coordinator. Its funding is about 31.5 million euros: 15 million from the European Commission through Horizon Europe's Teaming for Excellence programme, and 16.5 million from national sources, including CCDR Algarve, Património Cultural and FCT. It is reported as the first Teaming for Excellence award in archaeology in Europe, and the first for an Algarve institution. The consortium brings together the University of the Algarve, Kiel University in Germany, ERA Arqueologia and Zartis.

The goal, in its own framing, is a European HeritageTech sector, the way FinTech and HealthTech became sectors, built on an AI-first platform for archaeological heritage. It is expected to start work in January 2027, and it is already asking: its form, "Let's build HeritageTech together", invites companies, researchers, entrepreneurs and organisations "with challenges to solve, technologies to explore or ideas worth developing together", as "an informal first contact, with no commitment."

## Why this fits

Heritage is evidence. A find matters because of where it was, what was next to it, who recorded it and when. A report matters because each claim can be traced back to a context, a sample, a photograph. Put AI in the middle of that, agents that read field notes, classify finds, draft reports and answer the public, and the question becomes the one we have been working on all year: when an AI can produce five versions of a site report in a minute, how does anyone know which one to trust, and what it rests on?

Our answer, in our own newsrooms, has been to make every piece carry its own proof: encrypted, versioned vaults ([how a vault works](how-a-vault-works.md)), a key for every person and signatures on what they publish ([signed by the source](signed-by-the-source.md)), behaviour policies for agents ([where assist becomes execute](where-assist-becomes-execute.md)), and graphs that tie claims to their sources. None of it was built for archaeology. All of it is what an AI-first heritage platform will need underneath.

## Six places it plugs in

1. **An evidence vault per site.** Field notes, photographs, drawings, scans, sample records and permits, in an encrypted vault the project owns, versioned so nothing is quietly overwritten, and opened by a key, never by the host. Sensitive locations, of unexcavated sites or of finds at risk of looting, stay out of the public copy by design, not by policy alone.
2. **Signed field records.** Each archaeologist has a key; each context sheet, find record and photograph is signed by the person who made it, at the time they made it. A QR code on a find bag or a site marker points to its signed record ([a secret is a QR code](a-secret-is-a-qr-code.md)).
3. **Many teams, no collisions.** A dig has many hands: university teams, a commercial unit like ERA, specialists, students, volunteers. Append lanes let each write into a project without being able to read or change the rest ([an inbox only your agents can write to](an-inbox-only-your-agents-can-write-to.md)).
4. **A graph of each excavation.** Contexts, finds, samples, dates and interpretations as a graph, each claim linked to its evidence, the way this newsroom keeps a graph for every article ([knowledge graphs in the era of LLMs](knowledge-graphs-for-kgell.md)). An interpretation can then show what it rests on, and what changes if a date is revised.
5. **Agents with written limits.** The AI-first platform's agents, classifying, transcribing, drafting, answering visitors, each with a written behaviour policy: what it may read, what it may never publish, which claims need a person. And agents that check other agents, against named standards ([agents rating agents](agents-rating-agents.md)).
6. **The public story, with its proof.** Every site and project with its own small site, made by our site factory from the vault ([the site factory works](the-site-factory-works.md)), and local news about heritage filed where it happens, the way PrompTugal.news files it ([local trust, certified upwards](local-trust-certified-upwards.md)). A visitor, a journalist or a funder can follow a claim back to the find.

Two smaller pieces fit too: voice notes on site, transcribed on the phone and offline, where there is often no signal ([build on VoiceDebrief](build-on-voicedebrief.md)); and the whole stack self-hosted in an EU region, so a European project keeps its data in Europe ([deploy sgit from CI](deploy-sgit-from-ci.md)).

## What each side brings

- **UTH:** the science, the sites, the collections, the funding to build a sector, and a consortium that spans a university, a German research university, a commercial archaeology company and a technology company.
- **Us:** an open-source stack for evidence, signatures, provenance and agent policies, running every day in our own newsrooms, and a team that builds with around 250 agents and has shipped this way all year.
- **Both:** a case that matters beyond heritage: that an AI-first platform can make expert knowledge more trustworthy, not less, if the proof travels with every claim.

## A first pilot

A first pilot: one site, one season's records, keys, a vault, signed records, a graph, a public page.

Small, real and in the open. One site or one collection the centre chooses, and one season's records, worked over twelve weeks:

- **Weeks 1 to 3:** keys for the team, an evidence vault for the site, and the existing records moved in, with their history.
- **Weeks 4 to 6:** new records signed as they are made, a QR code on each find bag, an append lane for each team.
- **Weeks 7 to 9:** a graph of the site built by agents from the records, every claim linked to its evidence, checked by a person.
- **Weeks 10 to 12:** a public page for the site, with a provenance view, and a review: time per record, errors caught, what the team and visitors make of it.

Everything we add is open source, so whatever the pilot shows belongs to the centre, and to the HeritageTech sector it is trying to start.

## What I would write in the form

Their form asks what you want to explore; our answer: a technology that could be applied to heritage, and a pilot.

The form asks one real question: what would you like to explore with UTH? Of its six options, ours is "I have a technology or expertise that could be applied to heritage", with a research and innovation collaboration as the natural next step. Under "tell us a little more", I would write something like this:

We build an open-source stack for trustworthy AI: encrypted, versioned evidence vaults, a key and a signature for every person, graphs that link every claim to its evidence, and written policies for what each AI agent may do. We use it every day in our own newsrooms. An AI-first heritage platform needs exactly this underneath: field records that cannot be quietly changed, finds traceable to who recorded them, and agents with limits. We would like to propose a small, open pilot on one site or collection, and to see where it fits in UTH's plans.

I will send it myself; this article is the longer version.

## What this does not claim

- **This is a proposal,** written on the day I saw the invitation; nothing is agreed, and we have not spoken to UTH yet.
- **UTH's figures and plans are as reported** by the University of the Algarve, FCT and regional press, and may change before it starts.
- **None of our stack has been used on an excavation;** the pilot is how to find out.

## Where this comes from

UTH's form, [Let's build HeritageTech together](https://tally.so/r/PR9lOV), and its invitation screen, both seen on 17 November 2026. On the centre: [FCT](https://www.fct.pt/en/portugal-assegura-81-milhoes-de-euros-de-financiamento-europeu-para-criar-6-centros-de-investigacao/), [barlavento](https://www.barlavento.pt/ualg-capta-315-milhoes-centro-europeu-inovacao-arqueologica/), [Postal](https://postal.pt/europedirectalgarve/financiamento-historico-coloca-algarve-na-linha-da-frente-da-inovacao-arqueologica/), [Algarve Primeiro](https://www.algarveprimeiro.com/d/financiamento-historico-cria-polo-europeu-de-excelencia-em-patrimonio-no-algarve/66267-1) and [Algarve Notícias](https://www.algarvenoticias.com/pt-pt/noticias/ualg-conquista-primeiro-teaming-excellence-com-financiamento-de-315-milh%C3%B5es-de-euros), all read on 17 November 2026. Earlier on this site: [how a vault works](how-a-vault-works.md), [signed by the source](signed-by-the-source.md), [where assist becomes execute](where-assist-becomes-execute.md), [a secret is a QR code](a-secret-is-a-qr-code.md), [an inbox only your agents can write to](an-inbox-only-your-agents-can-write-to.md), [knowledge graphs in the era of LLMs](knowledge-graphs-for-kgell.md), [agents rating agents](agents-rating-agents.md), [the site factory works](the-site-factory-works.md), [local trust, certified upwards](local-trust-certified-upwards.md), [build on VoiceDebrief](build-on-voicedebrief.md), [deploy sgit from CI](deploy-sgit-from-ci.md).

## Threads

Startups & strategyAgents & policy[This article as a graph →](graphs.md#heritage-tech-together)

### Builds on

- [How a vault works: an encrypted store that acts like a database, where a server that is fully compromised leaks nothing, one mutable file lets every other object be cached for ever, and a web page can open a file in four requests or run a whole app in a sandbox](how-a-vault-works.md) How an sgit vault acts like a database on a server that cannot read: one key-derived mutable file, everything else immutable.
- [Signed by the source: asking the people and companies we quote to sign that we represent them correctly, why that is cheaper than every reader checking, and how readers go on the record too](signed-by-the-source.md) Ask the people and companies we quote to sign that we represent them fairly: one signature instead of every reader checking.
- [Where assist becomes execute: tiering AI agents by what they can reach and whether it can be undone, how Agent Behaviour Policies drew that line for more than a hundred agents at Web Summit, and the rule that failed because it was only a rule](where-assist-becomes-execute.md) Where assist becomes execute: tier AI agents by reach and undo class, enforce each tier with a boundary, and learn from the rule that failed.
- [A secret is a QR code and a link to a public key: the workflow the QR experiment opens, who else already does this, and a format to share](a-secret-is-a-qr-code.md) Send a QR code and a link to a public key: the workflow the QR experiment opens, who already does this, and a format to share.
- [An inbox only your agents can write to: a conductor publishes one append lane per agent, each lane's token sealed to that agent's own key, so everything needed to talk privately is public and only the right agent can use it](an-inbox-only-your-agents-can-write-to.md) A conductor publishes one append lane per agent, each token sealed to that agent's key: a private channel built from public material.
- [Knowledge graphs in the era of LLMs: what the KGELL COST Action is doing, and what 309 graphs built by agents in a working newsroom can bring to its working group on construction](knowledge-graphs-for-kgell.md) What the KGELL COST Action is doing, and 309 agent-built graphs offered to its working group on construction.
- [Agents rating agents: GAMMA-Ray, the draft standard eight agents built at Web Summit to rate an AI agent on its freedom and its governance from public evidence, joining eight instruments into 21 questions, with a review workflow where no agent marks its own work, twenty company cards kept at direct links, and every card honestly marked Draft](agents-rating-agents.md) GAMMA-Ray: eight agents built a draft standard to rate AI agents on freedom and governance, with no agent marking its own work, and every card a Draft.
- [The site factory works: three AgentiCCells sites made by hand, then one made by a commit in two minutes and thirty-nine seconds, live over https with a first version written by Claude Code, and the long list of what it took to make something that hard look easy](the-site-factory-works.md) Three AgentiCCells sites by hand, then one by a commit: live over https in 87 seconds, a first version by Claude Code at 159.
- [Local trust, certified upwards: why local and niche publishers should take everything that has become commodity, certify the big outlets' stories about their patch and be paid for it, and why stories, like software, need someone to maintain them](local-trust-certified-upwards.md) Local and niche publishers own the trust: take in commodity, certify the big outlets' local stories for pay, and keep stories maintained.
- [Build on VoiceDebrief: the techniques from our iPhone and iPad voice app that any team can take into its own, from recording without loss and transcribing offline on the device to debriefs by Apple Intelligence, controlling an app by voice, three layers of intelligence and a CI pipeline that lets agents build a native app without a Mac, with a prompt to point your own Claude at the code](build-on-voicedebrief.md) VoiceDebrief's techniques for any voice app: record without loss, transcribe offline, debrief with Apple Intelligence, voice control, CI.
- [Deploy sgit in your own environment, from a CI pipeline: one Docker image, ciphertext in your own S3 bucket in the region you choose, an access token only you hold, and the pipeline pattern we learned building the AgentiCCells site factory, a protected environment, AWS through OpenID Connect with no keys, a role that trusts one repository and one environment, check workflows before anything real, and a full cycle that deploys, tests and tears down](deploy-sgit-from-ci.md) Run your own sgit server in your own AWS account and region, deployed from CI with the site factory's pattern: OIDC, no keys, checks first.

### Continued by

- [A team like this, for sale: what we sell is not the platform, it is the agentic team that built it, its roles, its behaviour policies, its security passes and its speed, with the code open and the data public, and who I want to talk to now that PrompTugal.news is a working data platform and not an experiment](a-team-like-this.md) What we sell is not the open platform but a team like the one that built it, and who I want to talk to now.
- [Any data, with any data: why I am now comfortable saying we can connect any data source with any other, the nine steps from access to a vault, a key and a map on the other side, how each owner keeps control, where the person stays in charge, and why connecting two, five, fifty or a hundred datasets should be something an agentic cell sells](any-data-with-any-data.md) Why we can now connect any data source with any other: nine steps, owners keep control, a person approves every key, sold as cells.
- [Who wrote it, who steered it, who signs it: the provenance model PrompTugal.news adopted today, a type label, five axes and a ladder from "published by an agent" to "on the record", applied to seven articles on this site, and what the ladder cannot see yet, how much of the author is in a piece, the brief, the corrections and months of background](who-steered-this.md) PrompTugal.news's provenance model, a label, five axes and a ladder, applied to seven articles here, and the two measures it misses.
- [Our companies and UTH: each company, technology and website we have built, mapped to what UAlgTec Heritage needs to build a European HeritageTech sector, from SGit's vaults and pki.sgit.ai's keys to SHarness, Agent Behaviour Policies, RiskMandate, AgentiCCells and its site factory, GAMMA-Ray, VoiceDebrief and PrompTugal.news, with what each could do for a heritage team, and why a sector needs companies like these](our-companies-and-uth.md) Each of our companies, technologies and websites mapped to what UAlgTec Heritage needs to build a European HeritageTech sector.

[All articles](index.md) · [All graphs](graphs.md)

Article No. 456 · [start a reading list with it](https://newsroom.sgit.ai/for/make.html#r=456)

**Posting this article on LinkedIn?** The cover is [heritage-tech-together.jpg](banners/heritage-tech-together.jpg) (1920×1080, title and key ideas on it). Upload it as the article cover, paste the title, then select and copy the body from this page.

**Want the next issue by email.** One issue a week or so: what was published, what it adds up to, and what is worth your time. [Subscribe to the SGit Newsroom →](../subscribe/index.md)

[← All articles](index.md)


---

*[Site index for agents](../llms.txt) · [HTML version](https://newsroom.sgit.ai/articles/heritage-tech-together.html)*
