Vignesh | Product Designer

Go Back

Climaty AI - agentic

Designing an AI workspace for planning, launching, and managing campaigns

An agentic marketing platform that helps advertisers, agencies, and holding groups plan campaigns, collaborate with AI agents, monitor performance, and manage their marketing operations in one workspace.

My Role

Product Designer

Team

Product Manager

Engineering

Timeline

Jan 2026 – Apr 2026

tools

Figma, Claude

Project & Business Context

Climaty AI was evolving into an agentic marketing platform for teams managing campaigns at different levels of scale.

Users

Three types of user for this platform:

Advertisers – Teams managing campaigns and performance for their own brand.

Agencies – Teams managing campaigns across multiple clients and brands.
Holding Group – Larger organizations overseeing multiple agencies, brands, teams, and accounts.

MY ROLE

I worked with the Product Manager to understand the requirements, map the platform structure, and design the core workflows across campaign creation, performance monitoring, governance, and organization management.

The Problem

Marketing teams work across campaigns, brands, platforms, and teams. As the organization grows, these workflows become increasingly fragmented.

problem statement

How might we bring complex marketing workflows into one AI workspace without making the experience feel complex?

proof

Without a clear structure, users could lose context while switching between campaigns, AI agents, brands, platforms, and organization settings.

Research (Stakeholders / End User)

I worked with stakeholders to understand how different marketing organizations structure their work and manage campaigns.

what they said

Advertisers mainly work within their own brands.
Agencies frequently switch between clients and campaigns.
Holding groups need visibility across multiple organizations.
Users need clear context while moving between workflows.
Different AI capabilities need to feel like one connected experience.

Mapping the System First

Before designing screens, I mapped how the core parts of the product connected.

Different workflows need different experiences. Chat works best for active campaign creation, while Governance and Optimizer use dashboards for monitoring and ongoing actions.

Personal ≠ Organisation settings. Personal settings cover individual preferences and security, while Organisation settings manage brands, users, advertisers, billing, and access.

Quick Product Walkthrough

The Approach

I focused on making the product feel like one connected AI workspace instead of a collection of separate AI tools.

The structure keeps campaign and organization context consistent while users move across different workflows.

I focused on making the product feel like one connected AI workspace instead of a collection of separate AI tools.

The structure keeps campaign and organization context consistent while users move across different workflows.

Key Decisions

Connect specialist AI agents inside one campaign workflow instead of treating them as separate tools.

Connect specialist AI agents inside one campaign workflow instead of treating them as separate tools.

why

Users want to complete a campaign, not decide which AI agent to use. Connecting the agents keeps the workflow focused on the user's goal.

Multi-agent connections happen in chat. “Connect Multi Agents” sits beside Chat Home and Recent Chats, making agent work conversational rather than navigational.

Give Optimizer its own dashboard for ongoing campaign monitoring and insights.

Give Optimizer its own dashboard for ongoing campaign monitoring and insights.

why

Monitoring is different from active chat. A dashboard makes performance, trends, opportunities, and actions easier to scan.

Optimizer AI got its own dashboard. Performance monitoring is a glance-and-check habit, not a back-and-forth. Overview, insights, and actions all live in one view.

Give Governance its own dashboard for managing platforms, access, and compliance.

Give Governance its own dashboard for managing platforms, access, and compliance.

why

Governance requires ongoing visibility across platforms and accounts. A dedicated dashboard keeps these controls predictable and easy to manage.

Governance AI is a system of record, not a chat. It audits ad creative across Google Ads, Meta, TikTok, DV360, and CM360, with Pre, Post, and Ongoing types plus a searchable status table. Unlike chat, it keeps platform history easy to find.

Separate Personal and Organisation settings based on the jobs users need to complete.

Organisation settings mirror how agencies actually work. SSO, User, Advertiser, and Brand Management plus Billing all sit under Organisation, because team, client accounts, and billing are the same category of work, separate from personal settings

why

Personal settings focus on the individual, such as profile, security, notifications, and usage. Organisation settings handle users, advertisers, brands, billing, and account-level management. Separating them keeps both areas easier to understand and scale

Organisation settings mirror how agencies actually work. SSO, User, Advertiser, and Brand Management plus Billing all sit under Organisation, because team, client accounts, and billing are the same category of work, separate from personal settings

Constraints

Support advertisers, agencies, and holding groups within one product.
Handle different organization and account structures.
Maintain context across multiple specialist AI agents.
Balance chat-based workflows with dashboard-based monitoring.
Support multiple advertising platforms and integrations.
Keep the information architecture scalable as new capabilities were added.
Design while product requirements were still evolving.

Support advertisers, agencies, and holding groups within one product.
Handle different organization and account structures.
Maintain context across multiple specialist AI agents.
Balance chat-based workflows with dashboard-based monitoring.
Support multiple advertising platforms and integrations.
Keep the information architecture scalable as new capabilities were added.
Design while product requirements were still evolving.

Outcomes

81% active adoption across pilot marketing teams.

62% reduction in repetitive operational tasks across campaign workflows

3× faster campaign planning and optimization workflows
45% reduction in manual reporting effort
30+ workflows connected across planning, analytics, reporting, and governance
92% user confidence in AI-generated operational recommendations

Numbers tracked across pilot teams and campaign workflows during product testing.

What I Learnt

This project changed how I think about designing agentic products.

Not every AI workflow needs to be chat-based. Chat works well for active, goal-driven tasks, while dashboards are better for monitoring, comparison, and management.

Mapping the system first helped me choose the right interaction model for each job while keeping the overall product connected.

This project changed how I think about designing agentic products.

Not every AI workflow needs to be chat-based. Chat works well for active, goal-driven tasks, while dashboards are better for monitoring, comparison, and management.

Mapping the system first helped me choose the right interaction model for each job while keeping the overall product connected.

This project changed how I think about designing agentic products.

Not every AI workflow needs to be chat-based. Chat works well for active, goal-driven tasks, while dashboards are better for monitoring, comparison, and management.

Mapping the system first helped me choose the right interaction model for each job while keeping the overall product connected.

You can find me here :)

Instagram

You can find me here :)

Instagram

You can find me here :)

Instagram