Emilian is a driven health tech and longevity entrepreneur and investor, who transformed his own health crisis into an obsession to help people living healthier and longer. Despite past metabolic issues, at 50 years old, he is in the best health of his life. A decade ago, despite leading a healthy lifestyle, he experienced burnout and illness due to his hectic life and work. He analyzed his biomarkers, uncovering links between inflammation, hormones, and health. Through research and self-experimentation, he overcame his health issues. This sparked his mission to help others heal. He founded Ilara Health, a diagnostics company increasing healthcare accessibility in Africa, and Expand Health, a hybrid platform that offers a combination of technology and human expertise to enhance and convert current health and wellness spaces into longevity clinics, in South Africa, Eastern Europe and the Middle East. At 49, with biomarkers as proof, he feels better than ever. His goal is to empower 1 million health professionals and bring healthy living to billions.
Methodology:
- Downloaded data on 350 churned customers from HubSpot CRM into an Airtable, including info like customer ID, lifetime revenue, products purchased.
- Built a Glide application on top of the Airtable to create the interface for salespeople.
- Salespeople access the app before calls with churned customers.
- Call is recorded and transcription is generated via speech-to-text.
- App contains fields to capture notes, historical notes, AI-generated churn reasons categorized by themes, AI-suggested next steps for each reason category, key takeaways, and likelihood to return.
Tools:
- HubSpot CRM - for downloading customer data
- Airtable - to store and organize customer data
- Glide - to build application interface
- OpenAI integration - to generate text for churn reasons, next steps, takeaways based on call transcription input
Results:
- Provides clear and actionable churn reasons and recommendations without sales manager manually listening to calls
- Focuses attention on addressing issues to regain customers
- Can analyze conversation patterns over time with transcript records
Learnings:
- Technique could be applied in any sales business to surface insights from customer conversations and drive retention
- Could be adapted for other business uses like patient diagnosis and treatment planning based on doctor-patient conversations
Key advantage is using AI to quickly parse conversations to surface key information and recommendations. Main limitations may be quality/accuracy of the AI text generation (however it proved successful for broken English spoken in a rural medical setting in Kenya).
Methodology:
- Downloaded data on 350 churned customers from HubSpot CRM into an Airtable, including info like customer ID, lifetime revenue, products purchased.
- Built a Glide application on top of the Airtable to create the interface for salespeople.
- Salespeople access the app before calls with churned customers.
- Call is recorded and transcription is generated via speech-to-text.
- App contains fields to capture notes, historical notes, AI-generated churn reasons categorized by themes, AI-suggested next steps for each reason category, key takeaways, and likelihood to return.
Tools:
- HubSpot CRM - for downloading customer data
- Airtable - to store and organize customer data
- Glide - to build application interface
- OpenAI integration - to generate text for churn reasons, next steps, takeaways based on call transcription input
Results:
- Provides clear and actionable churn reasons and recommendations without sales manager manually listening to calls
- Focuses attention on addressing issues to regain customers
- Can analyze conversation patterns over time with transcript records
Learnings:
- Technique could be applied in any sales business to surface insights from customer conversations and drive retention
- Could be adapted for other business uses like patient diagnosis and treatment planning based on doctor-patient conversations
Key advantage is using AI to quickly parse conversations to surface key information and recommendations. Main limitations may be quality/accuracy of the AI text generation (however it proved successful for broken English spoken in a rural medical setting in Kenya).
