Bruno Paccini Cofounder & CEO
Was the first hire and CPO at Shine, a bank for small businesses in France, and led its product from inception through acquisition and scale.
The agent built for teamwork.
We’re delegating increasingly ambitious work to AI. As agents become capable of working for days, then weeks, the next step is delegating entire projects and long-term goals.
But the more ambitious the goal, the more collaborative the work becomes.
It requires expertise, context, and judgment from across a team. Decisions made by one person shape the work of another. Different contributions have to come together, often over weeks or months.
We believe that’s where the larger opportunity for AI lies.
Today, everyone directs their own agents, carries context between tools, reviews the work, and brings the results back to everyone else. As the amount of work agents can do grows, so does the work of coordinating it.
We believe AI can do more than make every person individually more productive. It can become part of how a team works together.
As agents become capable of doing more work for longer, they can take on the coordination too: carrying context between people, connecting work happening across a team, knowing when someone’s judgment is needed, and keeping things moving over time.
The consequence is much bigger than making every person more productive. It changes what a group of people can accomplish together.
Till is an agent built for teamwork.
You give Till a goal and work with it in one shared conversation.
People bring their expertise, judgment, and context. Till takes on work, coordinates it across people and agents, and brings people in when their contribution is needed.
Behind the scenes, Till has its own computer. It can work across the data, tools, and external agents each person gives it access to, while respecting the boundaries between them.
Till can also create and coordinate specialized agents to get things done. The team doesn’t have to manage how that work happens.
People shouldn’t have to learn the abstractions of AI to benefit from it. They should be able to work with agents the way they work with each other: through conversation.
Because Till is part of that conversation, it develops an understanding of how the team works. It sees the decisions people make, the expertise they bring, what works and what doesn’t, and the unwritten rules that shape how things get done.
Over time, Till can take on greater responsibility: following work as it evolves, connecting what happens across a team, and carrying projects forward over weeks and months.
We believe the next generation of companies will be shaped not only by how capable their people and agents are, but by how effectively they can work together.
That’s what we’re building Till for.
Was the first hire and CPO at Shine, a bank for small businesses in France, and led its product from inception through acquisition and scale.
Was CTO of Giskard, an AI security startup and research lab, where he worked on LLM safety and agent red-teaming. PhD from ENS Paris.
Prefer a conversation? Get in touch.