How Cloudflare’s APAC Team Booked 100 Demos in 2 Weeks With 2 Reps + AI Agents

Cloudflare’s APAC team wanted to generate more qualified pipeline without scaling SDR headcount or relying on generic high volume outbound.

We built an AI powered outbound system that continuously identified companies showing relevant infrastructure signals, researched why each account might need Cloudflare, and generated personalized outreach around that specific trigger.

With just 2 sales reps supported by AI agents, the system helped book 100 demos in 2 weeks.

Traditional outbound made it difficult to know which companies were actually worth contacting.

Sales reps were spending too much time:

• Building account lists
• Researching company infrastructure
• Looking for relevant buying signals
• Understanding what problem Cloudflare could solve
• Writing personalized messages manually

The bigger issue was timing.

A company changing its cloud infrastructure or launching a new website may suddenly have a strong reason to evaluate security, performance, CDN, or networking infrastructure.

But without continuously monitoring those events, reps often reached accounts too early, too late, or with messaging that had no connection to what the company was currently doing.

We wanted to build a system where the signal determined who entered outbound and what message they received.

Instead of building static prospect lists, we created five continuously running lead generation workflows.

We monitored companies for changes in their cloud and infrastructure stack.

The system identified signals such as companies adopting, replacing, or expanding technologies related to:

• Cloud providers
• CDN infrastructure
• Hosting
• Security
• Networking
• Web infrastructure

AI agents then researched the change and determined whether there was a relevant Cloudflare use case.

This gave reps a concrete reason to reach out instead of sending a generic infrastructure pitch.

We continuously tracked newly launched websites and digital properties across target markets.

A new website often creates immediate requirements around:

• Website performance
• CDN
• DDoS protection
• DNS
• Application security
• Infrastructure scalability

The system identified relevant companies, enriched the account and decision makers, and automatically added qualified opportunities into outbound campaigns.

This allowed the APAC team to reach companies while infrastructure decisions were actively being made.

We tracked LinkedIn posts from technical leaders and target accounts talking about cybersecurity issues or complaining about their current cloud provider, especially Vercel and AWS.

The system looked for posts discussing topics such as:

• Security incidents and breaches
• DDoS attacks and bot traffic
• Vercel bills, limits, or outages
• AWS costs, complexity, or outages
• Website performance and downtime
• Cloud migrations
• Infrastructure scaling
• CDN or networking challenges

Instead of simply treating engagement as a signal, AI agents analyzed the actual content of each post to understand the underlying problem.

Each post author's account was then qualified before it entered outbound:

• Company size and funding stage
• Current hosting and cloud stack
• Whether the problem mapped to a Cloudflare product
• Whether the author owned or influenced the infrastructure decision

Qualified accounts were then enriched and routed into the appropriate outbound workflow.

A company hiring for infrastructure or security is usually in the middle of fixing, scaling, or rebuilding something. We tracked new job openings for titles such as:

• DevOps Engineer
• Site Reliability Engineer (SRE)
• Platform Engineer
• Cloud Infrastructure Engineer
• Network Engineer
• Security Engineer
• Head of Infrastructure
• CISO

AI agents read each job description for mentions of CDN, WAF, DDoS protection, Zero Trust, or competing vendors like Akamai, Fastly, and Zscaler. A posting that named the exact problem Cloudflare solves moved the account to the top of the queue.

New technical leaders often review their vendor stack in their first 90 days. We tracked job changes and promotions into roles such as:

• CTO
• CISO
• VP of Engineering
• VP or Head of Infrastructure
• Head of Security
• Head of IT

When a security leader or CTO joined or was promoted at a target account, the system enriched their contact details and opened outreach around the priorities a new leader typically sets.

Finding the signal was only half of the system.

The next step was understanding why the signal mattered for that specific company.

For every qualified account, AI agents researched:

• What changed
• Why the company may have made the change
• Their current infrastructure
• Potential infrastructure limitations
• Relevant Cloudflare products
• The likely business or technical impact

Every first email was highly personalized to the signal that put the account into outbound: the stack change, the new website, the LinkedIn post, the job opening, or the new security leader.

AI agents also audited each account's website to qualify it and personalize the sequence:

• Traffic spikes and growth trends
• Website performance and page speed
• Security posture, including SSL, headers, and DDoS protection
• Gaps a specific Cloudflare product could close

Accounts with low traffic or strong existing protection were filtered out. For everyone else, short summaries of the traffic, performance, and security audits went into the follow-up emails, so each follow-up added something new and useful instead of just bumping the thread.

The goal was not to make emails look personalized.

The goal was to make the reason for reaching out genuinely relevant.

The final workflow looked like this:

1. 1Buying signal detected
2. 2Company researched
3. 3Traffic, performance & security audited
4. 4Account qualified
5. 5Decision makers identified
6. 6Contact data enriched
7. 7AI analyzes likely problem
8. 8Personalized messaging generated
9. 9Prospect added to outbound
10. 10Follow-ups share audit findings
11. 11Rep handles replies

AI agents handled the repetitive research and data work continuously.

The two sales reps focused primarily on conversations, qualification, and moving interested accounts into demos.

Every process in the system ran with Claude on a daily and weekly schedule, so outreach went out on autopilot while each signal was still fresh.

Daily runs caught new signals as soon as they appeared:

• LinkedIn posts about security issues and cloud complaints
• Newly launched websites
• New infrastructure and security job openings
• Research, enrichment, and messaging for new accounts

Weekly runs handled the slower-moving work:

• Cloud stack change scans
• Job changes and promotions of security leaders and CTOs
• Re-qualifying accounts already in the pipeline
• Reviewing which signals and messages earned replies

No rep had to remember to run anything. Accounts were reached early and on time, while the problem was still fresh.

Within two weeks, the system helped Cloudflare’s APAC team book 100 demos with just two sales reps.

Instead of asking reps to manually research hundreds of companies, the system continuously surfaced accounts with a reason to buy and gave reps the context needed to start relevant conversations.

The result was a fundamentally different outbound model:

The biggest change was not simply using AI to send more outbound.

It was using AI to decide who to contact, why now, and what problem to talk about.