# The Web Summit team, part two: seven people, each with their own agent, a designer with little time, and every voice memo turned into a follow-up within the hour, sgit.ai > The first brief for our Web Summit team set up an event site run by agents. This second part is about the people, because this is the first team we run with many humans in it. Seven of us are going, some in the Algarve and some still in the UK, all with ideas, and each of us brings our own agent, so the team is really a set of agents collaborating through their humans. That needs three things the agent-only teams did not. First, clear roles: who owns which company, who follows whom, who is talking to whom, recorded in a personal CRM for each of us, kept in our own vaults and synchronised into the team's view by shared ids. Second, speed: anything any of us records, a voice memo after a conversation, a scanned card, a note to our agent, is picked up within the hour and turned into action, a CRM entry, a page made for that company, a drafted follow-up, while a person still sends every email. Third, a way to work with people who have very little time, like the professional designer joining us, by learning what works for each person, briefing them in their own medium with the least context they need, and letting agents prepare the variants, the presentations and the animations around their work. And everything is worked backwards from the summit: the branded T-shirts, hoodies, cards and small things to hand out each have a date they must be ordered by, and each carries a code that leads to a page. Before the summit there are meetups in the Algarve; during it, the newsroom publishes; after it, the follow-up is the product we are showing. *Source: · 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) / The Web Summit team, part two: seven people, each with their own agent, a designer with little time, and every voice memo turned into a follow-up within the hour # The Web Summit team, part two: seven people, each with their own agent, a designer with little time, and every voice memo turned into a follow-up within the hour By [Dinis Cruz](https://sgit.ai/about/index.md) · 2026-10-31 · [article v1.0.0](versions/web-summit-humans-and-their-agents.md) · [newsroom v0.14.14](https://newsroom.sgit.ai/admin/versions.html) · web-summitagent-teamsbriefseventscrmdesigncollaborationarticle ***Abstract:** The first brief for our Web Summit team set up an event site run by agents. This second part is about the people, because this is the first team we run with many humans in it. Seven of us are going, some in the Algarve and some still in the UK, all with ideas, and each of us brings our own agent, so the team is really a set of agents collaborating through their humans. That needs three things the agent-only teams did not. First, clear roles: who owns which company, who follows whom, who is talking to whom, recorded in a personal CRM for each of us, kept in our own vaults and synchronised into the team's view by shared ids. Second, speed: anything any of us records, a voice memo after a conversation, a scanned card, a note to our agent, is picked up within the hour and turned into action, a CRM entry, a page made for that company, a drafted follow-up, while a person still sends every email. Third, a way to work with people who have very little time, like the professional designer joining us, by learning what works for each person, briefing them in their own medium with the least context they need, and letting agents prepare the variants, the presentations and the animations around their work. And everything is worked backwards from the summit: the branded T-shirts, hoodies, cards and small things to hand out each have a date they must be ordered by, and each carries a code that leads to a page. Before the summit there are meetups in the Algarve; during it, the newsroom publishes; after it, the follow-up is the product we are showing.* Seven people, each with their own agent, and the team's agents between them, on one event site. ## In short - **This is the first team we run with many humans.** Seven people going, each with ideas, each bringing their own agent. - **The team is agents collaborating through their humans.** Roles have to say who owns what and who follows whom. - **Everyone keeps a personal CRM in their own vault,** synchronised into the team's view by shared ids for every company and person. - **Anything recorded is actioned within the hour:** a memo, a scanned card, a note becomes a CRM entry, a page and a drafted follow-up. - **A person still sends every email.** The agents draft; nothing goes out on its own. - **People with little time get briefed in their own medium,** like the professional designer joining us, with the least context they need. - **Everything is worked backwards from the summit,** starting with what has to be ordered and printed. ## The first team with many humans The [first brief](web-summit-team-brief.md) for our Web Summit team set up the event site and the agents: a conductor, a researcher, a cartographer, a writer, a curator, an analyst, a designer and security, on the template the talk's, the book's and the cold read's teams used this week. Those teams were mostly agents, with me as the one human. The site itself is [web-summit.sgit.ai](https://web-summit.sgit.ai/), which has existed since 28 October 2026 with a team roster, its own agent and six vaults on our send server, public and private, for the CRM, the mail (Email-FS) and the task board (Issues-FS), each holding only a README so far. This one is different. Seven of us are going to Lisbon. Some live in the Algarve and some are still in the UK. All of us have ideas, and each of us brings our own agent, the way I work with mine. So the team is not a set of agents with one person on top. It is a set of agents collaborating through their humans, and the people need to be aligned before, during and after the summit as much as the agents do. ## Roles, and who follows whom With many people and many agents, the thing that breaks first is knowing who is doing what. So the roles are written down the same way an agent's are: - **Each company and each person on the map has one owner** among the seven of us, and the owner's agent keeps the record. - **Each of us has a published role** on the site: what we are following, which conversations we lead, and what we are not doing. - **The conductor routes,** as on every team: a request that names a company goes to its owner, and a request that names nobody goes to the board. - **Meetings and briefings are booked from the map,** so the person who walks into a meeting has the briefing their agent prepared, and the person who owns the company knows it happened. ## A CRM for each of us, kept in sync [The inbox team as a recipe](the-inbox-team-as-a-recipe.md) describes the CRM our email agents keep: one file per contact, thread and organisation, with a timeline that any agent may add to and none may rewrite. The Web Summit team uses the same shape, with one change: everyone has their own. - **The team's CRM vaults already exist** on web-summit.sgit.ai, one public and one private; they are where the shared view lives. - **A personal CRM in each person's own vault,** kept by their own agent, because each of us has our own relationships and our own email. - **Shared ids** for every company, person and event on the map, so two of us meeting the same company on the same day write to the same record. - **A team view on the site,** built from what each personal CRM chooses to share: who owns the company, the status, the next step, the pages made for them. What is private stays in the person's vault. - **A key for every agent,** each vouched for by its person, so the team can tell which agent wrote which record, as [agents have an identity now](agents-have-an-identity-now.md) describes. ## From a voice memo to a follow-up, within the hour Within the hour: a memo or a scanned card becomes a record, a page and a drafted follow-up that a person sends. We are selling an agentic workflow, so the way we follow up is the demonstration. The rule for the team is that anything any of us records is picked up and acted on within the hour, and often within minutes: 1. **Something is recorded.** A voice memo after a conversation, a photo of a business card, a note to our agent, a QR code scanned at our stand. 2. **The owner's agent picks it up** at its next wake, within the hour. 3. **The record is updated:** the person and company on the map, the conversation in the CRM timeline, the next step. 4. **An asset is made:** a page for that company, with what we discussed and what we can do for them, because behind every email there is an asset. 5. **A follow-up is drafted** in the person's own email, waiting for them. 6. **The person sends it.** No email goes out on its own; an email cannot be unsent. Done well, someone we met in the morning has a page made for them and a note from the person they met before the afternoon sessions start. ## A designer with very little time A professional designer we know is joining the team to help with the design. They have very little time, and they are a good example of the people we will need more of: experts whose hours are scarce, who should spend them on the part only they can do. So the job for the team is to make their hours count: - **Learn what works for them** first, and write it down as their card on the site: the medium they work in, how they like to receive a brief, how much context they want, how fast they reply. - **Brief in their medium.** If they work in a design tool, the brief arrives there; if they prefer a call, the agents prepare it and record the decisions. - **The least context they need:** one page on what we are, who it is for and what it must do, with examples of what we like and do not like. - **Agents do the rest around them:** the variants, the sizes, the print-ready files, the presentation that explains a design, the animation that shows it moving, so the designer decides and the agents produce. - **Short loops:** a few options, a decision, the next step, rather than long rounds. What we learn from working this way becomes a template for every scarce expert we work with after them. ## Work backwards from the summit I would like us to look like a team in Lisbon: branded T-shirts and hoodies, cards for everyone, small things to hand out, a banner. These have the hardest deadlines of anything we do, because they have to be designed, approved, ordered, printed and delivered before 9 November. So the plan works backwards: - **For each item, the date it must be ordered by,** from the printer's lead time, and the date its design must be approved, before that. - **Every printed item carries a code** that leads to a page: the person's card to their page and their agent, a flyer to the newsroom's summit site, so that what we hand out starts a follow-up. - **Who receives it, and where:** some of us are in the Algarve and some in the UK, so delivery addresses are part of the plan. - **What we can sell** at the summit, and how someone buys it from a phone. ## Before, during and after - **Before:** meetups in the Algarve with the startups and people we already know, the map filled, meetings booked, and the newsroom publishing about the summit and the companies going. - **During:** seven people, their agents and the team's agents; a daily debrief; the map updated after every meeting; content published as it happens. - **After:** every conversation followed up, every page made, and a debrief of the team itself. The follow-up is the product we are showing. ## What I need to decide - **Who owns which companies,** and each person's role card. - **The design budget,** and the order-by dates once we have quotes. - **How each person's agent joins:** its key, its vault and what its CRM shares with the team view. - **The designer's card,** once we have asked them how they like to work. ## What this does not claim - **It is a brief.** The personal CRMs, the shared ids and the hour-long loop are designed here; the email workflow and the CRM shape they build on run today at RiskMandate, and web-summit.sgit.ai's vaults exist but are empty. - **The people are not named.** The team's seven and the designer are described by role; each will choose what their card on the site says. - **The lead times are not known yet;** the dates come from the printers' quotes. ## Where this comes from - My voice memo of 31 October 2026, word for word in `admin/notes/2026-10-30__web-summit-team/memo-02.md` in this site's repository. - The Web Summit site's repository, [SGit-AI/SGit-AI__Website__Web-Summit](https://github.com/SGit-AI/SGit-AI__Website__Web-Summit), `REALITY.md` at site v0.1.9, read on 31 October 2026. - Related: [a brief for the Web Summit team](web-summit-team-brief.md), [Web Summit 2026, the plan](web-summit-2026-the-plan.md), [the inbox team as a recipe](the-inbox-team-as-a-recipe.md), [agents have an identity now](agents-have-an-identity-now.md). ## Threads Startups & strategyAgents & policy[This article as a graph →](graphs.md#web-summit-humans-and-their-agents) ### Builds on - [A brief for the Web Summit team: an event site run by an ephemeral agent team, built from this week's template, with a map of every company we should meet](web-summit-team-brief.md) A brief for the Web Summit team: an event site run by an ephemeral agent team, from a proven template, with a map of every company to meet. - [The inbox team as a recipe: how the RiskMandate agents run my email and CRM, written down by one of them so a second team can copy it](the-inbox-team-as-a-recipe.md) The RiskMandate inbox team, written down by its own agent so a second team can copy it. - [Agents have an identity now: a key pair made in every session, published with its role, so any agent can seal a secret to it, and the piece the cloud providers could add](agents-have-an-identity-now.md) A key pair made in every agent session, published with its role in a repository, so any agent can seal a secret to it; what the clouds could add. - [Web Summit 2026, the plan: a briefing for the team on usage first and the sale after, intelligence from public material, an asset, a page and a link for every person we want to reach, a newsroom of our own for the summit, and the two agentic workflows the team will learn on](web-summit-2026-the-plan.md) Web Summit 2026, the plan: usage first, an asset, a page and a link for every person, a summit newsroom, and two agentic workflows. [All articles](index.md) · [All graphs](graphs.md) Article No. 337 · [start a reading list with it](https://newsroom.sgit.ai/for/make.html#r=337) **Posting this article on LinkedIn?** The cover is [web-summit-humans-and-their-agents.jpg](banners/web-summit-humans-and-their-agents.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/web-summit-humans-and-their-agents.html)*