Wonderful episode! Would love a deeper dive into the development of the LLM model and how the copy output was ultimately inserted into outbound emails. Looking forward to part 2.
This is mind blowing. Honestly, it's painful to iterate on e-mail segmentation, without deep tech. As Emmy said, an army is necessary to make it valuable.
It would be amazing if in part 2 you shared some of the prompts, automations and workflows that allows you to come up with the summaries and recommendations for each new contact that entered your system.
Amazing episode! Thank you guys! Would love to understand more about the process of tuning (I guess some RAG happening behind the scene - how to feed behavioural data to the LLM etc. )and vector database search.
Ooooo that’s smart! I’m normally against AI-generated emails (especially for cold emails). But the idea of personalising emails to high volume inbound leads based on their inputs is clever!
Thanks for this episode! I'd love to learn more about how a marketer at a much smaller company than Hubspot could mirror this approach. What are the ways you could scale this down and still generate results at startups?
The LLM should have responded with, “hey guys, no personalization needed here, everyone selling cold brew has trouble selling it in winter. Don’t go thinking I reached AGI yet”
Wonderful episode! Would love a deeper dive into the development of the LLM model and how the copy output was ultimately inserted into outbound emails. Looking forward to part 2.
This is mind blowing. Honestly, it's painful to iterate on e-mail segmentation, without deep tech. As Emmy said, an army is necessary to make it valuable.
It would be amazing if in part 2 you shared some of the prompts, automations and workflows that allows you to come up with the summaries and recommendations for each new contact that entered your system.
Amazing episode! Thank you guys! Would love to understand more about the process of tuning (I guess some RAG happening behind the scene - how to feed behavioural data to the LLM etc. )and vector database search.
Ooooo that’s smart! I’m normally against AI-generated emails (especially for cold emails). But the idea of personalising emails to high volume inbound leads based on their inputs is clever!
❤ Great show!
Thanks for this episode! I'd love to learn more about how a marketer at a much smaller company than Hubspot could mirror this approach. What are the ways you could scale this down and still generate results at startups?
Thanks for bringing this out
This is really REALLY cool. Thanks for sharing.
How does the content get into the email? Is this done manually or with AI?
Share the graphic of the 2x2?
The LLM should have responded with, “hey guys, no personalization needed here, everyone selling cold brew has trouble selling it in winter. Don’t go thinking I reached AGI yet”
😂😂😂
this one
Learn how marketers use AI to personalize content with our FREE State of Marketing Reporting 💥 clickhubspot.com/j5t