A
antgoldbloom
I’m Anthony, co-founder/CEO of Sumble. I was previously co-founder/CEO of Kaggle. Sumble is my newco with Ben Hamner (former co-founder and CTO of Kaggle).
### What we built
Sumble is a knowledge graph for go-to-market teams. We allow you to run very rich queries to identify prospects at a granular level and be able to do very targeted outreach.
Sumble allows you to find:
- tech stacks (in larger companies, down to the team or buying group level) - key projects those teams are working on (cloud migrations, GenAI initiatives, etc.) - people involved in those key projects
For example, here's a list of GenAI projects at Capital One that involve RAG/Vector databases: Sumble
And this view includes a list of people who we think are involved in a particular project being undertaken by the AI Foundation Team at Capital One: Sumble
These views allow you to reach out to that team with a granular understanding of what they are working on.
### Inspiration
Sumble was very much inspired by our experience at Kaggle:
1. Kaggle’s public-data platform showed us how hungry people are for high-quality data (the metrics on that product were really strong)
2. At Google we saw knowledge graphs unlock powerful and composable queries
### Trying it out
Comments URL: Show HN: Sumble – knowledge graph for GTM data – query tech stack, key projects | Hacker News
Points: 68
# Comments: 45
Continue reading...
### What we built
Sumble is a knowledge graph for go-to-market teams. We allow you to run very rich queries to identify prospects at a granular level and be able to do very targeted outreach.
Sumble allows you to find:
- tech stacks (in larger companies, down to the team or buying group level) - key projects those teams are working on (cloud migrations, GenAI initiatives, etc.) - people involved in those key projects
For example, here's a list of GenAI projects at Capital One that involve RAG/Vector databases: Sumble
And this view includes a list of people who we think are involved in a particular project being undertaken by the AI Foundation Team at Capital One: Sumble
These views allow you to reach out to that team with a granular understanding of what they are working on.
### Inspiration
Sumble was very much inspired by our experience at Kaggle:
1. Kaggle’s public-data platform showed us how hungry people are for high-quality data (the metrics on that product were really strong)
2. At Google we saw knowledge graphs unlock powerful and composable queries
### Trying it out
- The app is live today; you’ll need to log in (Google OAuth or magic links)
- Most functionality and data are free; we only charge individual users for bulk exports
- Sources: job posts, resume data, company websites (more to come!)
- Extraction & linking: We use LLM (mostly fine-tuned models) to extract entities out of text from sources (company → team → people on a team → projects the team is undertaking → technology the team uses)
- Adding more sources so you can run even more composable queries
- Opening an API so devs can hit the graph directly
- Much later: expand to use cases beyond GTM
- Is the web app intuitive?
- What queries do you want us to prioritize supporting in an API?
- What additional external data sources would you like us to prioritize? - What workflow improvements/integrations would you find most helpful?
Comments URL: Show HN: Sumble – knowledge graph for GTM data – query tech stack, key projects | Hacker News
Points: 68
# Comments: 45
Continue reading...