Episodes
-
- About Hoxton
- Background at Google
- From California to London
- Early steps in data driven approach
- Being single person “data team”
- Designing and capturing signals
- Building ETL system
- Data challanges
- Data decay problem
- On ML stack
- ChatGPT impact on the process
- Team involvment with the tool
- The Rob score
- Usage across VC value chain
- Value of time series and following company progress over time
- Data stack
- Scraping tips to crawl anything
- How data approach changes VC industry
- Arms race of data
- Future of VC industry and founders raising money
- Importance of surfacing data
- VC as a sales job
- The perfect CRM tool
- How LLM’s change VC game
- Personal CRM as extension of mind
- Challange with follow-ups
- Advice for VC firm which start data approach
- Lack of tools in Europe
- Where to find Rob
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
Abel on social media:
Linkedin
Twitter
Today my guest is Abel Samot - VC Associate and head of data at Red River West. Abel has background in computer science and he has been passionate about bringing data approach to VC industry. At Red River West he has been at the forefront in building whole infrastructure from scratch and shaping data culture. Red River West has specific investment thesis and that shapes also needs for insights. In our conversation we talk about importance of data, legal aspects, how to shape data culture at the company, importance of UI and UX of a product and how LLM’s can potentially change the whole game for startups and investors.
p.s. Our conversation was recorded in September of 2023. Before launch of custom gpt’s and assistant api.
* Background of Abel
* Incremental journey to RAMP (Red River Algorithmic Management Platform)
* RRW investment thesis
* Using RAMP across VC pipeline
* Importance of quality of data
* Approximating company progress in different scores
* Cultural shift to data-driven
* Making sure people use platform daily
* Product management approach
* Importance of UI/UX of a platform
* Transparency of signals
* Building vs. buying solutions
* AI vs. KPI driven approach by Olivier Huez
* Collecting data for future use
* Tech stack at RRW
* Legal aspects of data
* Stage of Data-driven industry
* How data-driven approach could shape future of venture capital
* 3 person unicorn startup
* Potential problem to differentiate yourself when everyone has data
* Advice for vc's who want to start data-driven approach
* Data-Driven VC: how data can help source & screen startups?
* Data-Driven VC: what we have built at Red River West
* The biggest challenges for data-driven VCs
* 10 ways to leverage LLMs as a Data-Driven VC
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
Missing episodes?
-
Today my guest is Sarah Guemouri - Principal at Atomico. She is responsible for data-driven initiatives like research, sourcing, investment opportunities and portfolio work. In our conversation we cover all of these aspects, deep dive in tech stack behind Atomico's "Boar" platform and discuss the ways how to shape data-driven culture. Plus some practical hacks. Also Sarah is one of the people who kickstarted Datahunt community for data people at VC firms.
Find Sarah online:
Atomico
Linkedin
Twitter
Notes
* Background and road to Atomico and AppAnnie (Data.ai) experience
* Data value chain at Atomico and how it was started
* From consulting approach to automation
* Internal startup at VC firm
* Boar platform
* Shaping data-driven culture
* Real world problems, incentives and psychology
* Practical culture hacks - leaderboard and communicating wins
* Building own tools vs. buying existing ones
* Tech stack development at Atomico
* LLM use cases and learnings
* Reconciliation layer on top of data
* State of data driven VC industry
* Building data-driven VC community
* On VC industry disruption
* Importance of human relationships
* Hands on advice for small VC's who want to be data-driven
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
Other platforms:
* Google Podcasts
* Apple podcasts
* Spotify
Today my guest is Andre Retterath - Partner at Early Bird Venture Capital and founder of Data-driven VC Newsletter. If you are listening to this podcast and haven't signed up already, you probably should do it. Andre is one of the pioneers in data-driven investing approach, have done academic research in the field and also he is hands an practitioner.
In our conversation we talk about his academic journey, augmented VC approach and where does gut feel fits in, VC-founder fit and how data is changing way both sides find each other. And also Andre gives some practical tips for those who want to start their data-driven VC journey.
Notes
- Background and how Andre got into VC industry
- On intersection between practical and academic sides
- Why VC's are not more data-driven - culture, data
- Gut feeling, biases and dependence on network
- State of data-driven vc industry
- First wave of big data in vc industry and why it died
- Data-driven landscape insights
- VC value chain
- Who are driving change in VC industry
- Investment thesis - hard, soft criteria and alignment of partners
- Concept of augmented VC approach
- It's not only about finding the deal but also getting into the deal
- Founders don't want to work with algorithm
- Room for gut feel
- Augmented VC article
- Early bird Eagle eye investment platform
- Information democratisation through chat interface
- Using OpenAI and dealing with sensitivity of data
- Big public models vs. smaller internal models
- AI impact on productivity
- Thoughts on agents
- Data-driven investing from founder side
- Future of finding VC-founder fit
- Thesis matchmaking
- Data-driven innovations in VC industry
- Practical tips for VC firms who want to start data-driven approach
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
Other platforms:
* Google Podcasts
* Apple podcasts
* Spotify
Show notes
- Pietro’s background and journey from consulting to EQT Labs
- Building digital culture at EQT
- Three generations of Motherbrain and way ahead
- Potential for data-driven private equity
- Fishing vs. hunting. Noticing company vs. finding one.
- Implementing digital transformation at EQT
- Importance of having someone at the top who is excited about the change
- ChatGPT revolution
- Three stages of AI awareness
- Illusions of AI strategy
- Formless.ai, Intercom.
- Sam Altmen congress hearing quote - AI doesn't perform jobs, it performs tasks
- Validating mental model against ChatGPT
- Second and third order effects of AI.
- AI first thinking in every industry and AI driven bakery
- Million idea anxiety
- Concept of craftsmanship developers
- Kevin Kelly 1000 true fans
- Zed IDE
- AI first experience
- Emergence of agents
- Challenges with AI tools
- Hugging Face, Prefect. Code-less functions
- In-house models vs. “big ones”.
- Helm model benchmark
- Vibe test of models
- Anthropic 100k context model
- GPT4 explaining GPT2
- Model psychology
- Model alignment and right prompting
- Anxiety about AI disrupting companies
- Working on 10 year bets
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
This is the second episode in our data-driven VC’s series. Mike is co-founder and Applied AI at Moonfire - one of the pioneering VC firms in the field of data-driven investing. Mike is a hacker at his heart and we go into some interesting technical details in our conversation.
You can listen to Starwatcher podcast over here or pick you favourite platform:
* Google Podcasts
* Apple podcasts
* Spotify
Show notes and topics we are discussing:
- Mike's background as a hacker, engineer and becoming a VC
- Establishing Moonfire with data as a core fabric of the firm
- Sourcing, screening and evaluation as a machine learning driven recommendation problem
- Adopting transformer architecture from very beginning.
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
- Universal approximation theorem
- About the stage of Data-driven investing industry
- Data-driven VC landscape report
- Analysing founders vs. analysing companies. Units of calculation at different stages of company life cycle.
- How LLM's have affected people analytics
- Importance of data across sourcing, screening and evaluation cycle
- Thesis-driven approach
- Having prepared mind in conversations with entrepreneurs
- Importance of identifying which companies don't fit thesis.
- Trusting data vs. making an own opinion
- Assisted expert decision making process
- Building tools to assist human expert
- Agent based approach
- Impact of public Large language models
- Story about improving venture scale classifier model by 20%
- Performance of GPT4 vs. in-house models
- Kardashev scale
- API (outsource) models vs in-house models
- Giving LangChain access to internal tools
- Diminishing AI advantage in VC industry
- AutoGPT and "Thought, Action, Observation" loop
- Alignment problem
- Why it is actually valuable to be kind with LLM's
- Dramatic change of intellectual labor market
- Importance of humans to humans experience
- AI from startup perspective
- Rapid change of information exchange and expressing human emotions
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com -
This is the first episode in our data-driven VC series. Amir is a Senior Vice President of Fund & Investment Analytics at Techstars and has ton of knowledge. Hope you enjoy our conversation. Let us know what are your thought in comments.
Perspectives shared are Amir's and not those of Techstars
Show notes and topics we are discussing:
* Role of the data person in the company and also in context of portfolio
* Signal to noise ration across investment scene
* Being a data detective. Everyone has the same data but different conclusions
* Data roles in data organisation and team work
* State of Data-driven investing
* About getting the right data from companies
* Products which generate unique data not available anywhere else and ARK Kapital
* SVB case - signal vs noise.
* Jesse Livermore on fear and greed
* Market shocks and change of culture
* Court of Versailles Vs. The Wild West
* Risk reward trade-off in VC industry
* Astrology of VC industry
* Data-driven vs. So what?
* Large language models
* Embeddings as a way to represent information
* Agents and agent based models
* Hallucinating and improving prompts
* Blind reliance on tools and FOMO
* Potential biases in the data and how that affects results
* Language models from startup side of the table
* BabyAgi
* Dramatron
* Cereal box pitch
* SVB Risk assessment by ChatGPT
* Information arbitrage and obligation to invest
* Money Ball vs Babe Ruth
* Valuation distribution and risk reward dynamics
* Becoming managers of agents
* Exponential curve of innovation - The AI Revolution: The Road to Superintelligence
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit starwatcher.substack.com
