When we build AI solutions for clients, we focus on the intelligence—the part that helps people work smarter, move faster, or communicate better.
But smart agents need more than brains. They need a place to live.
They need an interface. A memory. A way to talk to each other.
That’s why we built the Activate Bar: a collection of modular tools that handle all of that.
It’s the infrastructure behind our AI solutions—designed to help us move fast, launch faster, and scale smarter.
Whether we’re working with a corporate comms team, an innovation lead, or an agency partner, the Bar gives us and our clients a head start.
🧩 Built for Different Needs
- For comms and PR teams, it means tools that reduce repetition, keep track of conversations, and let AI handle the busywork—without needing to learn a new system.
- For innovation leads, it means a flexible platform that integrates easily and shows results quickly—without needing to pull in a full dev team.
- For agencies and partners, it means a ready-made stack that can be customised, branded, and reused across clients—no need to reinvent the wheel.
- For product builders, it’s a shortcut to a stable, scalable backend, with orchestration, interfaces, and collaboration built in from day one.
🍸 What’s in the Bar
Every bar has its essentials—and ours is no different.
The Activate Bar includes a set of tools, each designed to do one thing well.
They can be served neat, or as a flight—depending on what your team needs.
🍊 Spritz
The Orchestrator
Spritz connects all agents and tools into one cohesive system. It routes tasks, manages execution, and keeps everything in sync.
What it’s good for:
- Handling multiple agents in complex workflows
- Making the system modular and easy to extend
- Letting agents discover and delegate to each other
Used in:
- Daily briefings for public affairs clients (with multiple agents summarising, filtering, and formatting info)
- Agent-to-agent workflows like: fetch amendments → extract text → send to analyst → generate email summary
- White-label deployments where agents must behave differently based on organisation or channel
🍸 Gimlet
The Document Builder
A structured writing environment that uses blocks and templates to assemble smart documents collaboratively.
What it’s good for:
- Creating standardised outputs like newsletters, reports, and press releases
- Embedding agent-generated content inside templates
- Letting humans and AI co-write in a predictable format
Used in:
- Commission reports where each section is filled by a different agent
- Automated press release drafts, editable by clients
- Building briefing documents with version history and reusable templates
🥃 Old Fashioned
The Messaging Hub
A central inbox that receives messages from multiple channels and sends them to the right agent, keeping context intact.
What it’s good for:
- Letting clients interact with agents from tools they already use (Slack, email, WhatsApp)
- Managing long or asynchronous conversations
- Avoiding context loss across systems or threads
Used in:
- White-label email inboxes where clients can “ask” for daily updates or summaries
- Slack integrations where users can drop a link and ask for a briefing
- Hybrid workflows where a human and an agent both contribute to a conversation
🍹 Daiquiri
The Chat Interface
A clean, conversational UI for talking to agents. It handles natural queries, triggers tasks, and keeps everything in threaded context.
What it’s good for:
- Quick, informal interactions with agents
- Letting users explore what the agents can do
- Tracking multiple tasks in parallel
Used in:
- Internal agent testing and demos
- Lightweight customer-facing chat tools for white-label deployments
- Prototyping agent interactions without needing a full UI
🧉 Mojito
The Editor
A minimal, collaborative writing space with real-time AI suggestions, editing tools, and version tracking.
What it’s good for:
- Improving and rewriting content with agent support
- Human-in-the-loop editorial work
- Templated but flexible writing
Used in:
- Press release review where the agent proposes improved copy
- Translation workflows with suggestions shown side-by-side
- Editorial “finishing” of agent-generated briefings
⚙️ Why It Matters
The Activate Bar lets us skip the boilerplate and focus on what actually matters: building useful agents.
Under the hood, it’s powered by a modular architecture where each agent acts as an independent micro-service.
Agents declare their capabilities, respond to structured tasks, and can delegate work to one another.
This makes the system flexible, composable, and easy to extend.
- New agents can be added without touching the rest of the stack.
- Existing ones can be updated or swapped out without disruption.
- We can plug into existing systems—or launch entirely new ones—without starting from scratch.
That’s how we move from prototype to production in days, not months.
And it’s why we can deliver white-label products that are fast, reliable, and shaped around what our clients actually need.