Easy Navigation

The Challenge
What I did
My Solution
Impact

The Context

The startup was in an early growth phase, setting up its operations while scaling across multiple brands.
“Break things fast” was part of the culture. Experimenting, shipping, and learning rapidly. When more systems and brands came online, it became clear that speed needed a stronger foundation to hold up over time.

The Core Problem

The business was moving quickly, and the foundations were still being put in place. There was no central CRM, Slack had become the default place for questions and decisions, processes weren’t clearly documented, and project management tools weren’t yet supporting the way work actually happened.
On the execution side, lead conversations were handled by AI, but the underlying qualification and follow-up logic wasn’t structured or reusable across brands.

This led to:

• Reporting that was fragmented and manually maintained
• Fragmented customer and sales data
• Repeated questions and decisions flowed through Slack, so the founder became the bottleneck
• Work only moved with constant manual oversight; without it, progress stalled
• Significant time spent on low-leverage tasks like content publishing and admin

The business continued to hire as it grew, but without strong foundations, each new role added coordination and oversight rather than reducing load.

The Systems I Designed

Rather than adding more tools or automations in isolation, I designed an AI-enabled operations layer that embedded decision logic, removed manual effort, and allowed the business to scale without increasing dependency on people. The focus was on structure first, automation second.

1. Decision-Led AI Messaging & Engagement System

Lead conversations were already automated with an in-house AI-driven Instagram messaging system. But the underlying logic wasn’t structured or reusable. I worked closely with the head of sales to extract their sales methodology from their head: how they qualified leads, handled objections, and decided next actions and translated it into a structured, reusable decision framework.

This framework:

• Defined qualification logic, conversation flow, and escalation rules
• Standardised tone, intent, and progression across brands
• Removed reliance on the sales team.

The system dynamically shifted states:

• From Sales Mode → Customer Mode
• Loading different prompts, tone guidelines, and onboarding logic once a lead converted.

This allowed AI to handle conversations end-to-end while maintaining consistency and intent as volume increased.

Once a lead converted, the system automatically switched modes:

• From Sales Mode → Customer Mode
• Dynamically loading new system prompts, tone guidelines, and onboarding logic

This ensured customers were onboarded correctly without manual hand-offs. Over time, we continuously analysed performance data, tested variations, and refined the system, eventually consolidating everything into a reusable playbook the team could rely on.

2. Automated Content Publishing System

Content publishing was consuming significant time across multiple brands. A necessary but low-leverage activity.

I designed an automated content system that:

• Handled end-to-end publishing across five brands with multiple content formats.
• Removed manual scheduling and posting
• Maintained consistency without increasing workload

Once stable, I documented the system and created clear SOPs so the team could safely update prompts and extend the system without breaking it.

3. Core Operations, Speed-to-Lead & Visibility

To reduce reliance on manual oversight, I designed supporting operational systems:

• Real-time notifications into Slack when leads from a cold email campaign replied, calls were booked, or sales occurred. This eliminated delays and reduced the need for manual checking.
• Automated reporting to replace manual spreadsheet tracking with real-time reporting so sales and call data flowed automatically. I later worked with the ClickUp dev to get it integrated directly into ClickUp, which gave the entire team live visibility without reconciliation or admin.
• Set up an in-house AI-driven outbound email system designed to remove repetitive admin and scale outreach volume without additional headcount
• A central CRM to act as the source of truth for leads and customers
• Automated client onboarding to remove repetitive setup work

This meant that data flowed cleanly across systems, reduced duplicated data entry and gave the team shared visibility without constant checking or follow-ups.

The Resulting Operating Model

The outcome wasn’t just automation or using AI for novelty; it was a reliable operating model.

• Decisions no longer depended on individuals
• Significant reduction in manual workload
• Systems absorbed growth without becoming fragile
• Faster response times without increasing headcount
• Consistent sales and onboarding experience across multiple brands

This foundation allowed the business to continue experimenting and growing, without everything breaking underneath.