I was tasked with researching and designing a conceptual experience for an AI-driven Email Engagement SaaS tool that could be tested with users. The objective was to explore how AI could help users generate professional, personalised emails tailored to their goals and preferred tone—streamlining outreach while maintaining authenticity.

I started by speaking to several salespeople who regularly conducted email outreach. Using the "Jobs to be Done" method, I gained a clear understanding of their goals and functional needs for successful outreach. The key needs that emerged were:

1. Personalising emails, such as using names and finding common ground to build rapport and generate interest.
2. The ability to edit pre-generated emails.
3. Control over the tone of the email, ensuring it sounded human and not robotic.
4. Engaging subject lines to prevent emails from being ignored.
5. Including a clear call to action to encourage enagement and drive progress in the sales funnel.

To better understand the landscape, I researched existing products and tools in the email engagement space. The key questions I aimed to answer were:

1. How are these products or tools being used to meet user needs?
2. What functionality have users come to expect from these tools?
3. What are the limitiations or pain points user experience, and where are there opportuities for improvement?

Research highlighted two key factors for a best-in-class experience:

1. Low Interaction Cost: Most AI tools, like ChatGPT and Grammarly, required users to spend time refining prompts and repeating the process for every email. Our tool needed to simplify this by reducing manual effort.
2. Personalisation: Generic AI-generated emails lacked the recipient-specific context needed for engagement. With access to rich data, our tool could deliver highly personalised emails effortlessly, offering a clear competitive advantage.

These insights gave an early, clear picture of what the user experience needed to deliver. They also became a key reference point for senior management to understand user needs and guide the product’s strategic direction.

A key design focus was enabling the capture of high-quality data to fine-tune our AI models for generating relevant, personalised emails. Collaborating closely with a small team of AI engineers, I identified a company’s website as the most valuable data source. Using web scraping techniques, we extracted and fed structured information into a large language model (LLM).

This approach allowed the AI to develop a comprehensive understanding of the user’s company, including the industries it operated in, products and services offered, key use cases, target customer profiles, and even the most relevant people within companies to engage with—based on job title and seniority.

Put simply, by ingesting this data and leveraging AI, the tool could identify and prioritise companies most likely to buy a user’s product or service, suggest the best contacts to engage with, and generate highly personalised emails—a game changer.

To gather the data needed to train the language model, we integrated an optional step into the onboarding workflow. This approach was intentional for several reasons:

1. User Expectation: Users naturally expect to provide information during onboarding. The design required only three optional fields, making the process feel familiar and low-effort.
2. Faster AI Training: Capturing data early allowed us to start training the model immediately, significantly reducing the time it took to generate personalised emails.
3. Lower Interaction Cost: By collecting this data upfront, we removed the need for users to input it later. This reduced friction and helped them reach the product’s "Aha" moment—sending their first AI-generated email—more quickly.

This design choice balanced ease of use with technical requirements, ensuring a smoother onboarding experience while accelerating the AI’s ability to deliver value.

Working closely with a small team across Product, Engineering, AI, and Data, I defined a simple use case and a set of functional requirements to shape the engagement generation UI. This concept served as a starting point, bringing the idea to life while generating excitement and discussion.

The design helped align everyone—from squad members to senior leadership—by providing a clear understanding of the use case, user workflow, and overall vision. It also surfaced many key questions we needed to address.

If you’ve made it this far, thank you! I genuinely appreciate you taking the time to read about my work.

It was great working on this project—I see it as a strong example of how design can move quickly and "scrappily," while still maintaining a methodical approach. With a well-defined company strategy already in place, exploration and conceptual design played a key role in shaping how we could execute effectively to achieve it.