Zailor

Rethinking Revenue Workflows with Embedded Agents.

Role

Product Designer

Team

1 Designer (me), 1 Project Manager

Disciplines

Interface design, Prototyping, Product strategy, Research

Year

2025

How Might Embedded Agents Rethink Revenue Workflows?

Overview - In the summer of 2025, I led product design on an account tracker and GTM intelligence tool built for revenue and customer-facing teams. My focus was on interface design and research, exploring how embedded agents can transform static workflows into adaptive systems.

Revenue teams are buried in complexity.

Problem - SDRs, account executives, customer success managers, and revenue operations teams move between dozens of tools every day. They jump between enrichment platforms, CRM dashboards, outbound tools, LinkedIn tabs, internal spreadsheets, and shared documents. They search for leads, gather context, update records, and attempt to stitch all of it into a coherent motion.

Data can become fragmented. Insights arrive too late. Valuable time is spent managing systems instead of managing relationships.

Optimizing revenue workflows through embedded AI agents.

Solution - Zailor is designed to sit at the center of revenue operations. It is built on three core pillars: account tracking, embedded intelligence, and agent-driven orchestration.

Stay in context while you work

AI searches across the web to surface the latest company news, funding updates, and key data points, keeping everything you need in a single view with no tab switching.

Reveal enriched company insights

Access accurate headcount, annual revenue range, and key metrics at a glance, helping teams qualify accounts faster and focus on the opportunities that matter.

Build and deploy intelligent agents

Design adaptive agents that track accounts, surface insights, and move work forward automatically. Configure their behavior and action, starting from presets or building custom agents that match your team's workflow.

Learning Revenue Ops Through Collaboration

Research - To understand how revenue teams actually work, I conducted several co-design sessions with SDRs, AEs, and RevOps professionals. Rather than relying on paper flows or static wireframes, we built and iterated real flows together in Lovable, adjusting interactions live as they provided feedback.

Designing with the intended user base instead of for them allowed us to move faster, validate assumptions immediately, and ensure each workflow decision aligned with how they already think and operate. These sessions became the foundation for the final interaction model.

Research 1
Research 2

Exploring agent-powered flows through rapid, coded experimentation.

Prototyping and Testing - Instead of relying on static screens, I built interactive prototypes using Lovable, Next.js, and Cursor, which made trial and error faster and far more intuitive. Coding early meant I could test responsiveness, refine micro-interactions, and consider edge cases like loading and empty states from the start.

We then tested these flows with revenue professionals. Observing how they used them revealed opportunities to refine the experience.

Intelligent systems should remove friction, not control.

Reflection - Complexity should not always be hidden, it should be handled. This project showed me how intelligent automation can reduce the burden of tools without reducing capability. Designing with revenue teams kept the system grounded in reality, and building in code kept assumptions honest. I learned that the real challenge in agent design is balancing autonomy with trust, and ensuring technology acts on behalf of people, not the other way around.

Zailor Kanban
Zailor Chat
Zailor Agent 2
Zailor Agent 1
Projects

Zailor

Rethinking Revenue Workflows with Embedded Agents.

RoleProduct Designer
Team1 Designer (me), 1 Project Manager
DisciplinesInterface design, prototyping, product strategy, research
Year2025

How Might Embedded Agents Rethink Revenue Workflows?

Overview - In the summer of 2025, I led product design on an account tracker and GTM intelligence tool built for revenue and customer-facing teams. My focus was on interface design and research, exploring how embedded agents can transform static workflows into adaptive systems.

Revenue teams are buried in complexity.

Problem - SDRs, account executives, customer success managers, and revenue operations teams move between dozens of tools every day. They jump between enrichment platforms, CRM dashboards, outbound tools, LinkedIn tabs, internal spreadsheets, and shared documents. They search for leads, gather context, update records, and attempt to stitch all of it into a coherent motion.

Data can become fragmented. Insights arrive too late. Valuable time is spent managing systems instead of managing relationships.

Instead of forcing simplicity, we can use AI as a layer that optimizes complexity.

Opportunity - Simplifying this complexity by removing features isn't the answer as it limits capability. The real challenge is to embrace that complexity and design workflows that align with individuals and teams, not against them. Instead of shrinking the toolset, we can connect it. Instead of relying on people to push data forward, we can embed AI agents that act autonomously to reduce friction.

Optimizing revenue workflows through embedded AI agents

Solution - Zailor is designed to sit at the center of revenue operations. It is built on three core pillars: account tracking, embedded intelligence, and agent-driven orchestration.

Stay in context while you work

AI searches across the web to surface the latest company news, funding updates, and key data points, keeping everything you need in a single view with no tab switching.

Reveal enriched company insights

Access accurate headcount, annual revenue range, and key metrics at a glance, helping teams qualify accounts faster and focus on the opportunities that matter.

Build and deploy intelligent agents

Design adaptive agents that track accounts, surface insights, and move work forward automatically. Configure their behavior and action, starting from presets or building custom agents that match your team's workflow.

Learning Revenue Ops Through Collaboration

Research - To understand how revenue teams actually work, I conducted several co-design sessions with SDRs, AEs, and RevOps professionals. Rather than relying on paper flows or static wireframes, we built and iterated real flows together in Lovable, adjusting interactions live as they provided feedback.

Designing with the intended user base instead of for them allowed us to move faster, validate assumptions immediately, and ensure each workflow decision aligned with how they already think and operate. These sessions became the foundation for the final interaction model.

Research Image 1
Research Image 2

Exploring agent-powered flows through rapid, coded experimentation.

Prototyping and Testing - Instead of relying on static screens, I built interactive prototypes using Lovable, Next.js, and Cursor, which made trial and error faster and far more intuitive. Coding early meant I could test responsiveness, refine micro-interactions, and consider edge cases like loading and empty states from the start.

We then tested these flows with revenue professionals. Observing how they used them revealed opportunities to refine the experience.

Intelligent systems should remove friction, not control.

Reflection - Complexity should not always be hidden, it should be handled. This project showed me how intelligent automation can reduce the burden of tools without reducing capability. Designing with revenue teams kept the system grounded in reality, and building in code kept assumptions honest. I learned that the real challenge in agent design is balancing autonomy with trust, and ensuring technology acts on behalf of people, not the other way around.

This project is ongoing.