The AI workforce 2 partly shipped · 5 building

The team a paid creator could never afford — shipping as part of the product.

A community manager, marketing expert, analyst, personalisation engine, setup assistant, plus AI search and AI-driven moderation. Two are partly shipped today; the rest build across the next six months. All run on the same unified fan graph — which is why Concord can ship them as a coherent team and a bolt-on AI tool cannot.

Partly shipped

AI Search foundation is in production and the AI Moderator's content-safety layer is enforced on every message today. The agent-style versions are next.

Shipping in months

Community Manager, Marketing Expert, Analyst, Personalisation Engine and Setup Assistant. All on the roadmap funded by this round.

The workforce

Seven AI workers, all running on the same unified fan graph.

Partly shipped

AI Search

Find anyone on Concord by what they mean, not what they typed.

Semantic search across the fan graph. A creator (or a fan) describes who they're looking for in natural language and Concord ranks results by meaning, with explanations of why each match showed up. The first AI capability shipped — and the foundation other AI workers build on.

Capabilities

Semantic search over every profile with synonym expansion
Match explanations so creators understand why a result showed up
Today: people search; building: creators, communities, events and paid content
Search-driven recommendations for fans and creators alike

Today vs. next

Semantic people search is in production. Expansion into creator, community, event and paid-content search is on the roadmap.

Partly shipped

AI Moderator

Reads every message and enforces policy at scale, with context.

Real-time content safety screens every message and image today, with per-space policies and audit trails. The next milestone is an AI moderator that understands conversation context — tone, dispute history, intent — instead of just keywords, and that can de-escalate without escalating to the creator.

Capabilities

Real-time content safety on every message and upload
Per-space content policies and allowlists configured by the creator
Audit logs and override tools for every action
Building: context-aware moderation that reads conversations, not just keywords
Building: de-escalation prompts and warning workflows before suspensions

Today vs. next

Real-time content safety is enforced on every production message and upload. The context-aware AI moderator agent is building.

Building

AI Community Manager

The 24/7 community manager a paid creator could never afford to hire.

An always-on community manager that welcomes every new member, answers the routine questions, walks new subscribers through their tier perks and escalates anything that actually needs the creator's attention. Reads every channel; remembers every fan; speaks in the creator's voice.

Capabilities

Personalised welcome flows for every new member and subscriber
Real-time FAQ answers drawn from the creator's pinned content and prior messages
Tier-perk onboarding walkthroughs for new paying members
Issue detection: flags drama, complaints and at-risk supporters before they leave
Smart escalation: only forwards the cases that genuinely need a human

Today vs. next

Foundations in place: full fan graph, real-time channel events, push notification infrastructure. The always-on agent loop is building.

Building

AI Marketing Expert

Plans launches, drafts the copy and sends it personalised — at audience scale.

Plans launches, picks the right time per fan segment, drafts announcements in the creator's voice and sends mass DMs or email campaigns where every fan gets a message written for them specifically. Tests, learns and recommends what to run next.

Capabilities

Launch planning with offer suggestions tuned to the creator's fan base
Announcement drafting in the creator's voice (multi-tone, multi-format)
Mass DMs and email campaigns personalised fan-by-fan, not segment-by-segment
Send-time optimisation per fan timezone and engagement window
A/B testing and post-launch analysis baked in

Today vs. next

Foundations in place: message infrastructure, push, full fan behaviour data. The marketing-expert agent loop is building.

Building

AI Analyst

Plain-English answers to the questions every creator avoids asking.

An always-on analyst that watches the numbers and answers questions the creator hates having to dig for: who's about to churn, what's underpriced, which space drives the most retention, which fan is ready to upgrade, what changed this week and why.

Capabilities

Plain-English Q&A over the full creator dataset
Weekly summaries of what changed, what worked and what to fix
Churn-risk scoring at the individual fan level
Upgrade-likelihood scoring and proactive expansion suggestions
Anomaly detection: flags when a metric moves before the creator notices

Today vs. next

Foundations in place: full analytics pipeline, event capture, notification trails. The analyst agent and its question/answer surface are building.

Building

AI Personalisation

Every fan gets a message written for them, at the scale of the whole audience.

Cross-cuts every other AI worker. Where the marketing expert plans the campaign and the community manager handles routine ops, the personalisation engine writes each individual message so every fan gets something that actually feels written for them — DMs, announcements, emails, gifts, check-ins.

Capabilities

Per-fan message generation tuned to the creator's voice and the fan's history
Personalised welcome flows for each new subscriber
Birthday notes, thank-yous, churn check-ins, upgrade nudges — every fan, every time
Crystal gifting suggestions tuned to each fan's behaviour
Personalised mass email campaigns at full audience scale

Today vs. next

Foundations in place: full per-fan event data and message infrastructure. The personalisation engine and creator-voice tuning loop are building.

Building

AI Setup Assistant

Tell Concord what you want. It builds it.

Removes the largest single switching cost for creators: re-setup. Describe the community, paywall, launch or migration in one sentence and Concord configures the spaces, roles, paywalls, schedules and notification flows behind it.

Capabilities

Natural-language community and paywall setup
One-sentence launch flows (spaces, paywalls, schedule, announcement, email all built)
Discord import driven by intent: 'bring my server, lock #premium behind a $9/mo tier'
Smart defaults for new creators based on their stated niche and goal
Re-runs and edits in plain English instead of menus

Today vs. next

Foundations in place: every underlying Concord API the assistant needs to call. The natural-language interface and planning loop are building.

Why this is defensible

AI tools without the fan graph are demos. The fan graph without AI is a database.

The fan graph is the moat

Every AI worker depends on knowing the full state of the creator's business: payments, roles, messages, interests, events, presence. Concord captures all of that from day one in one data layer. A bolt-on AI tool sitting outside the platform only sees fragments.

Each worker compounds the next

The analyst spots a churn risk. The marketing expert plans the save. The personalisation engine writes the individual message. The community manager handles the conversation. None of them work in isolation — they share a single context.

Premium AI is a real revenue stream

Creators already pay for tools that automate one slice of this work. Concord ships the whole team as a premium add-on with platform-grade margins. A standalone AI tool can be cloned in weeks; an AI workforce wired into the entire product stack cannot.

Round funds the rest of the team

$500K pre-seed. Ships the AI workforce that makes Concord uncopyable.

The data layer underneath is already running. Two AI capabilities are partly shipped. The next five workers — community manager, marketing expert, analyst, personalisation engine and setup assistant — ship across the next six months. The round is sized to fund creator distribution alongside this build.

invest@concord.digital Read the full deck Back to overview