Led the 0 to 1 of Nexthink’s first low-code orchestration product, evolving the platform from passive monitoring into an active, enterprise-grade remediation engine. Owned the full design lifecycle from discovery to launch, working in close partnership with 1 Product Manager and 8 Engineers to define the product vision, interaction model, and experience.

Impact

• Successfully launched and scaled the product to 300+ enterprise customers post-release.
• One of the fastest-adopted products in Nexthink’s history.
• Enabled a $20M cost avoidance for a single enterprise customer.
• Achieved a 99% success rate in remediating critical VPN failures.
• Drove Nexthink’s 0 to 1 transition from analytics only to an automation platform

Context: Historically, Nexthink focused on analytics and telling IT teams what was wrong. The business opportunity was to close the loop. We needed to move from diagnostics to resolution.

Problem: IT teams were overwhelmed with repetitive tickets (e.g. VPN drops, password resets). Existing tools were polarised, too technical, and required senior engineers to write PowerShell scripts. Second, it is too rigid and offering only pre-defined actions with no flexibility.

To ensure we solved the right strategy, I led a discovery phase involving 30+ interviews and monthly validation sessions with our enterprise partners.

We analysed five distinct users, ranging from IT Directors to L2 Engineers, to design a single platform that could satisfy their opposing technical needs.

Key Insights

• Routine tickets required agents to toggle between 4 disconnected tools, resulting in inefficiency. We identified the opportunity to centralise the entire resolution workflow into one place.
• The market was polarised between scripts and rigid actions. We targeted a low-code platform that combined the flexibility of scripting with an easy-to-use and accessible interface.
• Automation skills were locked away with senior engineers. The design needed to allow experts to build safe logic blocks that junior agents could assemble without breaking production.

During the early definition phase, we faced a critical architectural conflict. We initially explored a form-based approach because it was faster to build and aligned with standard IT form patterns. However, my prototyping workshops revealed a critical flaw. As workflows grew complex, the linear form collapsed under its own complexity, and users couldn’t visualise the logic flow.

Workshops: I facilitated a “Design vs. Dev” workshop to scope the MVP. We agreed to utilise the ReactFlow library to handle the canvas interactions, allowing us to ship the superior UX without scope creep.

Result: The node-based model aligned perfectly with the IT mental model of troubleshooting trees, creating immediate trust with our technical user base.

To ensure the platform could scale from 5 to 30+ integrations without creating a design bottleneck, I architected the “Thinklet” framework.

Instead of bespoke nodes for every action, Thinklets are atomic, reusable logic Nexthink blocks with standardised:

• Inputs & Outputs, creating a predictable data flow for users.
• Ensuring consistent behaviour across all integrations.
• Configuration, allowing Engineering to rapidly build connectors (ServiceNow, Jira, Teams) using pre-defined specs, without requiring new UI work for every release.

The prototyping logic wasn’t robust enough to simulate state loops and variables. To validate the core interaction model, I stepped outside of design tools and, working with developers, built functional prototypes using ReactFlow. This allowed us to test the actual limits of the engineering library and validate complex logic with users before development began.

Before committing to full development, I established a feedback loop with 8 enterprise customers to test the interaction model. We moved from static screens to code-based prototypes to validate complex logic.

Thinklets connections
Early testing revealed that as workflows grew, users struggled to distinguish between “Success” paths and “Error” paths visually.

Fix: I introduced a strict visual language for connections: Solid green for success and dashed red for failure/timeouts. This reduced configuration errors by making the logic flow scannable at a glance.

Troubleshooting workflows
Users reported that the results page offered no insight into why a workflow failed or at which step the logic broke.

Fix: I introduced an execution timeline that showed the exact path taken on the device. This allowed users to step through the execution, inspecting the specific inputs and outputs of each “Thinklet” to pinpoint errors immediately.

Flow successfully transformed Nexthink from a passive analytics tool into an active remediation platform. It became the fastest adopted product in company history, empowering IT teams to shift from reactive fixes to proactive automation.

• Flow successfully remediated a VPN client issue affecting 90% of their workforce.
• One customer used Flow to detect and reclaim unused Java licenses, resulting in $20M in cost avoidance.
• 35% reduction in support tickets by automating common fixes, customers significantly reduced their volume of recurring L1/L2 tickets.
• On average, customers save nearly 4,000 hours annually per workflow. One client saved $175k in just 12 hours by automating their SCCM maintenance.