The 102nd episode of Datacast is my conversation with Astasia Myers - a Partner on Quiet Capital's enterprise team leading investments in ML, data infrastructure, open-source, developer tools, and security. She focuses on pre-seed, seed, and Series A.

Our wide-ranging conversation touches on her upbringing in Silicon Valley, her education at Stanford and Cambridge, her early career in sell-side equity research, her time at Cisco’s corporate development team, her transition to venture capital at Redpoint, her current journey as a Founding Partner at Quiet; tactical discussions on early-stage hiring, product-led growth, community-led sales, angel investing; her enthusiasm for data-centric ML; and much more.

Please enjoy my conversation with Astasia!

Listen to the show on (1) Spotify, (2) Apple Podcasts, (3) iHeartRadio, (4) RadioPublic, (5) TuneIn, and (6) Stitcher

Key Takeaways

Here are the highlights from my conversation with Astasia:

On Growing Up in Silicon Valley

I grew up in Palo Alto and was exposed to technology and startups very early. My mother and our family friends all worked in tech. My mom worked in the semiconductor industry, truly the original Silicon Valley. I am lucky to have had her as a mentor growing up. I always found it exciting, especially enterprise technology, since the infrastructure was the foundation of other services.

Back then, it was a much smaller community that was more concentrated in the Bay Area. I would get to see a lot of the gadgets early before they would eventually go mainstream. I remember my mom worked on a semiconductor that was eventually used in the Sony Robotics Dog that was all over the news during that era. That was pretty cool, as it infused me with an appreciation for how that could be transformational and groundbreaking.

On Studying International Relations at Stanford

I got the deep tech from my mom and a penchant for travel from my dad. He was a commercial airplane pilot and always flew places. A lot of the time, I felt like growing up at the airport to pick him up from work. We went on many family trips together. I was humbled even to live abroad when I was growing up. My mom helped manage some teams in Denmark, so we spent time there. I did homestays in Japan because I took Japanese for a decade. In college, I even lived in Mongolia for some research projects. I have always just loved travel and experiencing other cultures. Understanding the relationship between technology and foreign cultures was a cool fusion of two passion areas.

I was humbled to have attended Stanford. It was a magical experience for me. Many diverse perspectives and passionate individuals surrounded me. Stanford really fosters an environment of exploration and creativity. During my time there, I wanted to fuse my passion for tech and international relations, so I researched the role of information technology, economic development, and politics. For example, that time was the era of the Arab Spring and the Iranian Green movement, and my work led to the creation of Stanford's Program on Liberation Technology, which is still a program today.

On Conducting Research with Professor Condoleezza Rice

I was fortunate to get to know Condi when she returned to Stanford after her stint as the US's Secretary of State. It was her first year back, and she actually taught a course on international policy - which I applied to and happened to get in. We built a relationship over a few years, and she was very supportive of students exploring their passions.

The research we conducted centered around the idea of political risk - the probability that a political action could significantly impact a company's business. Recall that this is the era when Google had just pulled out of China. Local acts were pushing back globalization. More businesses were operating in foreign frontier markets. Given my passion for studying how technology affects societies, it was interesting to reframe that work in terms of how political actors affected technology companies abroad. This could be an array of actors: Twitter users, local officials, or even hackers. 

Hydraulic fracturing was on the rise around the world. It is a technique for the oil and gas industry that includes injecting water, sand, and chemicals under high pressure into the bedrock to access the oil. It was slightly controversial because of the potential environmental damage. You can imagine that some tension arises due to this technique. Our research was around the actors, policies, and laws that affected businesses doing hydraulic fracturing. The study was eventually published by the Harvard Business School Press and used in Condy's political risk course at Stanford GSB.

On Studying Technology Policy in the UK

Growing up in Silicon Valley and attending Stanford, I wanted more exposure to how other communities thought about tech. I felt like I was in a tech bubble, so I wanted to study abroad. Cambridge, also known as Silicon Fen, was the center of tech in the UK at the time. Microsoft had research institutions there, as well as other industry-sponsored centers. Cambridge's engineering school is world-class in the area, consistently receiving some of the most venture capital money for early-stage startups. It was also a cool time when the Raspberry Pi computer started coming out of Cambridge. For me, it felt like an innovation hub.

The program I enrolled in allowed me to study how foreign governments thought about tech policy. How do you stimulate entrepreneurship? How do you support entrepreneurs in their journeys? How do you construct policies that allow foreign investment? That was really nice on the policy side, but I also simultaneously took courses at the engineering school. It was a nice program that allowed me to sit across two different institutions and gain an eye-opening experience on how the UK (and the broader Europe) thought about technology.

Silicon Valley was, and still is, a very mature ecosystem. There were repeat founders who had exited their businesses and gone to join venture capital firms or become angel investors. This idea of creating a business and receiving support from institutions or the community was mainstream. The government had traditionally supported academic research with financial grants. When I was in the UK, there was also this history of government grants and support of innovative research across STEM subjects. However, the ecosystem around starting and building a company was more nascent. Culturally, there was a different level of risk appetite. Additionally, the legal regime made it harder for founders to start businesses with more paperwork. It was interesting for me to be on the ground there and talk to people who are creative about starting businesses yet facing roadblocks in their way.

On Working as An Equity Research Analyst at Baird and Co.

I had been conducting research since high school and did it for nearly a decade. I personally enjoy distilling information into a clear thesis and moving forward with conviction. I knew I wanted to continue conducting research that was focused on technology. I thought the sell-side equity research would be a great fit because it combined my passion for tech with the business context, which I got from Cambridge’s business school.

While at Baird, I covered publicly traded tech companies focused on enterprise, specifically IT networking, security, and cloud businesses like Cisco, Palo Alto Networks, and VMware, among others. The work was really analytical and allowed me the opportunity to become an expert — resembling what I was doing in academia. It also honed my ability to evaluate a company and make a call on whether or not the business was fairly valued, overvalued, or undervalued. It was a pretty cool role because it encouraged me to think about

1. what was going on in the broader ecosystem outside of just the company I was covering,

2. what was going to disrupt the business, and

3. what was the new technology innovation or go-to-market motion on the horizon.

At the time, I could do deep dives on evolving technologies like software-defined networking and storage, which was a huge movement away from hardware, appliance-based architecture to software running on commoditized hardware. I always liked thinking about trends and trying to predict the future using data.

I will never forget covering VMware and discovering Docker, which was released within a year. It was a huge threat to virtual machines. I went to some meetups to learn more about Docker. During one of VMware’s analyst days, where the room was filled with hundreds of people, the CEO was onstage and asked the audience to raise their hands if they had heard of Docker. I put my hand up, one of only 3–4 people in this room. I realized at that moment my passion for sleuthing out new technology and going the distance to learn more about what is next. I knew I had to go earlier in the tech company’s lifecycle rather than focusing on publicly traded companies.

The sell-side equity research is a wonderful foundation. It allows you to have domain expertise. For me, that expertise was enterprise infrastructure and security. It teaches you ways to evaluate a business, both with financial skills and the understanding of management teams — their aptitude for the technologies at the individual companies and the broader ecosystem of alternatives. That helps you figure out the differentiation and trends these technologies are enabling. Finally, it creates an opportunity to think about disruption from both a business model and a technology perspective. I found that I love thinking about what is next/what is coming/what is on the horizon. Thus, I naturally gravitated to focusing on companies that were earlier in their development.

On Driving Venture Investments at Cisco

I covered Cisco at Baird, so I knew the company and its product lines really well. We used to do buyer surveys, and people consistently ranked Cisco at the top across multiple categories. I really admired the executive team and the business they had built. Cisco was one of the most acquisitive tech companies of all time and well-regarded for executing/integrating acquisitions. It was considered a superpower of the business. We used to joke that, sometimes, you could think of M&A as R&D for certain businesses. Cisco was at the forefront of thinking about the future. 

Knowing that I wanted to do more hands-on work with startups, I reached out to the head of corporate development at the time and eventually joined to work with the data center networking team, famously led by the MPLS team. It was pretty neat. I had a sense of where the views were going from my research. I felt like I could hit the ground running.

Cisco is so special because it is super creative. They would use any tool to add value, from M&A to investing to joint ventures:

• I got to work on pretty cool stealth projects that became product lines. I also invested in companies like Cohesity, Datos IO, Elastifile, and Guardicore.

• Cisco was known for spin-ins - where you essentially invest in a business and then have a call option to buy it at specific price points based on milestones achieved. I worked on one of those businesses called Springpath.

I cannot say enough good things about my experience at Cisco. I worked with incredible people. It is a special place that is supportive of employees and their next steps. It has a truly incredible track record of other people from the corporate development team going into venture investing - Mike Volpi (Index Ventures), Arif Janmohamed (Lightspeed), Max Gazor (CRV Ventures), Bucky Moore (Kleiner Perkins), etc. It is a great location where you can learn, grow, and find your path.

On Different M&A Frameworks

What is interesting about acquisitions is that: the framework the acquirer uses to determine the value of a business depends on the stage of the company. I like to think about it as three different buckets:

1. The first bucket is acqui-hire: These are people who are gifted in technology whom you are hoping to add to your team. They may have built some technology but do not have a solid go-to-market motion or have not exactly found a product-market fit. Those businesses are often valued based on the number of engineers on the staff. Think about it as very advanced recruiting.

2. The second bucket is companies with some traction and customers in the $1-10M ARR range. They are valued in the context of publicly traded companies, precedent transactions, and the immediate top-line value they could add to your business.

3. The third bucket is later-stage acquisitions. These are businesses with 10+M in revenue, and they could immediately be a standalone business unit. Cisco did that with acquisitions like Meraki or Jasper so that these businesses could continue their amazing growth trajectories. Those were most closely valued in terms of publicly traded comps.

That was cool to see different ways companies were valued. At the end of the day, it is all a negotiation. It was fun for me to work on some of those large acquisitions when I was at Cisco.

On Joining Redpoint Ventures

I really enjoyed my time at Cisco, which gravitated towards early-stage venture investing. I wanted to work closely with founders to help them achieve their dreams. The corp dev role at Cisco varied between companies. We worked on investing, M&A, and sometimes partnerships with stealth incubation. While I enjoyed the work, I really wanted to focus on venture. I worked on large acquisition opportunities that were considered transformational. Those were all hands on deck that could take you out of the venture market for six months, which is a very long time. So I naturally gravitated to wanting to work with early-stage founders and be an enterprise venture capitalist.

I became aware of Redpoint because they were an early investor in Springpath, a business I had worked with while at Cisco. They had super-deep domain expertise in enterprise technology, with investments in Juniper Networks, Hashicorp, Twilio, etc. Since enterprise was my passion, I was super excited to work with others in a domain I knew inside and out.

On Misconceptions About The Venture Industry

A silly one people believe is that we tweet all day. We actually do real work. We do a lot of deep research, spend time with founders, and help our companies. It is not just funny memes. Trust me.

Venture, as a financial instrument, has evolved a lot. The speed of innovation, similar to tech companies, has accelerated over the past five years. Even the world of venture at the time I joined Redpoint was very different from the world of venture now. 

• It has become more of a well-known asset class.

• There has become further specialization in terms of domain expertise and stage. 

• The expectations around what a business needs to prove before each stage of investment (seed -> A -> B) have transformed over the past five years. When I focused on enterprise software at Redpoint, a good business had achieved over 1M in ARR and a 3x growth trajectory. That is not necessarily how we think about a high-flying business today. Often, the business is earlier in revenue, but the growth trajectory is higher than 3x (which is qualified by their sales pipeline).

If I were someone thinking about entering venture today, I would clearly do my research to get a sense of the different stages of venture and know what makes me happy. Early-stage venture is very much a founder- and dream-oriented approach to technology, as compared to 5 years ago - which was having a product and customers to be onboarded. If you love data and financial analysis, you are probably more oriented to being an early-growth investor (Series B and beyond). But if you like being a thought partner for founders, noodling on the product, and helping with early recruiting/customer development, you are probably an early-stage investor.

On Forming Her Investment Theses

I was lucky because I came from Cisco, an infrastructure and security business, and was in sell-side equity research covering those categories before that. After being in this domain for a few years, I had a good sense of the publicly traded companies and startups disrupting them - particularly domains such as networking, databases, and dev tools. I felt lucky that I was coming in with an understanding of the market.

There are several ways people can get up to speed about a category:

1. Read press releases and news from TechCrunch, Silicon Angle, or third-party research reports from Forrester, Gartner, IDC, etc.

2. Talk about macro-trends and companies that operate in those spaces.

3. Check out blog posts from companies you are excited about or newsletters from well-regarded thought leaders (like Pete Soderling of Data Council).

The best way to get up to speed on categories is just reading, reading, and reading. By collecting all this information, you can distill it into your perspective of emerging categories and emerging tech trends.

On an average day, I spend about 1.5-2 hours sleuthing the Internet and trying to find cool blog posts, announcements of product releases, or open-source projects that have come to fruition. When doing due diligence on a company, I am the kind of person who drops everything and does deep, focused work. The due diligence process can be 10 to 20 hours in a very short period of time, weeks or even days - trying to understand a business. If you are thinking about going into venture, hopefully, you like to read.

On Her Investments In Solo.io and LaunchDarkly

As early-stage investors, we look for incredible founders who have a unique insight and vision for the future, a track record of execution, and very strong domain expertise in their categories. Both of these companies have incredible technical founders - Idit Levine at Solo.io and Edith Harbaugh and John Kodumal at LaunchDarkly. With both of these, it was quite apparent that the founders knew infrastructure inside and out.

For me, it always starts with the people. Our position is unique; we are supposed to be more than just financial partners but operating partners to a business. That comes through deep relationships with people to help them on their journey and be all hands on deck to help them succeed. Both of these companies had very compelling leaders.

Another factor that was pretty neat about these two was:

1. In the case of Solo.io, they had a cool architectural innovation. It is an API gateway and service mesh startup. In terms of the API gateway, they were plugging into an open-source project called Envoy, which came out of Lyft and brought about a unique approach.

2. In the case of LaunchDarkly, they had a feature management solution that allowed teams to deploy code more quickly in a secure manner. The founders had seen the value of feature flagging at their respective businesses. A build-or-buy decision was often made, generally applicable to a broad audience. Why not build a best-of-breed solution that enables people to do feature flagging effectively, so they do not need to take internal engineering resources to build it themselves? They were very early market entrants into a big space.

The third factor that got us excited was the large markets:

1. If you think about the history of API gateways, you had publicly-traded companies like F5 or an alternative like Kong that was valued at over 1B$. So we got excited about Solo.io because of its unique technology advantage in a huge category.

2. For LaunchDarkly, every single developer can benefit from feature flagging, which accelerates development velocity. They had a wide range of users - from publicly traded companies to even a dairy farm in Europe. An early-stage business can rarely have such a range of users benefitting from the product.

To summarize: Phenomenal technical leaders with unique insights + unique technology differentiation + a very large market they could address. As an early-stage investor, I have to put more weight on the team component. People with incredible technical skills, unique insight, and extreme drive often find a way to hire/recruit, build compelling products, and even create markets.

On Her Investments In Hex and Preset

Redpoint was an early investor in Snowflake, which ended up being the largest enterprise IPO ever. Watching that journey, we started to see the emergence of the modern data stack, enabling analytics and data exploration teams to answer tough questions. It is a fusion of data pipelining technologies (like Airbyte and Fivetran), data warehouses (like Snowflake, BigQuery, or Redshift), metrics stores that uniformly define metrics (like dbt Metrics or Transform Data), and a transformation layer that allows analytics engineers to transform data and make it usable for end applications. In terms of Hex and Preset, we had this vision that the interface for dashboarding and exploratory analytics would evolve, so we thought they fit well into this modern data stack architecture.

Preset is an open-source BI and analytics solution.

1. It was founded by Max Beauchemin, the creator of Superset, who decided to build a company around it. That is a great founder-market fit.

2. It had an interesting architecture because data engineers and data platform professionals adopted it. We thought it could have grown within organizations.

3. In terms of market, BI and analytics are one of the largest enterprise software markets in the world. We have many examples of large outcomes - from Google buying Looker to Salesforce buying Tableau.

In terms of Hex, I was spending a lot of time thinking through the evolution of notebooks.

• What I discovered there was that there were two different notebook audiences. There are data analysts or data professionals who really know SQL and are graduating in Python. Then, there are the deep data scientists and ML engineers who live and breathe Python and R, building models to be deployed into production. For the latter category, notebook solutions include Anaconda, Databricks, and AWS SageMaker. For Hex, their interface could also be applied to data analysts to make them more effective with exploratory data analysis.

• Hex really stood out to me because it fused a SQL editor with a notebook environment, bringing the ability to build data applications for less technical people to explore data and create data artifacts. Putting this together, Hex created a data workspace for teams. Seeing the implicated value of notebooks for their audience was very cool.

• Hex's founders were incredible. They had worked together at Palantir.

We also thought it was a new trend of the bifurcation of responsibilities between a dashboard and exploratory analysis - with the dashboard being Present and the analysis being Hex. It could be a huge market that they could go after.

On Advice To Her Portfolio Companies

The most important area for startups to focus on in the early stages is hiring and recruiting. It is a tougher climate to recruit engineers than ever before. Teams are being thoughtful and more open to hybrid and remote work. You can see that being reflected in the success of companies like Deel, which enables payroll for remote employees. Often these relationships with early employees are built over time. Here are my three pieces of advice:

1. Be open-minded to hybrid and remote work. There are incredible people all over the world who want to be part of a startup journey and are eager to add value.

2. Identify companies that have technical talent who could be valuable for your team. Try to find people who have been at those businesses for over two years. Reach out to them in a very customized context and talks about their unique skillset - how they would be special for your team. Get to know them over time to build a relationship so that you are top of mind if they consider a new opportunity.

3. Spend at least 30% of your time focused on hiring in the early days since it is so crucial to the success of the business.

It is also important to have a general sense of your go-to-market motion when you start your business. Traditionally, there has been tops-down, which is selling to an executive and going through a traditional sales process. This is also transforming into a bottoms-up motion where the individual makes a singular purchasing decision usually after. They have had a free trial or lightweight engagement with the product. These two different go-to-market motions affect the sales cycle, the demonstration of value, and how you build your product. It is essential to speak to your users, your stakeholders of the product, and the buyers early in the customer discovery process to identify which go-to-market motion could be best for you.

On Writing Comprehensive Primers

As you can tell, I have always enjoyed research and writing. It allows me to distill my thoughts and express myself to a broader audience so they can get a sense of my analytical mind. My process for writing these primer articles looks like this:

1. I first start engaging with founders. I speak to all of them to get a better appreciation of the teams, their thoughts on how their market is evolving, and the core tech differentiator of their product. You can often see pockets of innovation that emerge at similar time horizons. If I see a collection of new technologies trying to solve a similar pain point, I start to take note of it and do a market map of every relevant company in the space.

2. I then speak with buyers and operators of this potential new technology. For example, data orchestration technology is adopted by data engineers or data platform leaders. This helps me get a sense of what they value in technology, the ROI of adopting new tech, and what they would need to see in a product to make a buying decision.

Fusing these two audience insights with my own, I highlight key themes: Why is there a market shift? What is the underpinning trend? What would cause a user to pick one solution over the other? Then I tried to make a prediction around what I think will be the winning platform or the evolution of the space.

The writing process is not short. It often takes me about a month or so to distill all the information, crystallize it, and publish it. But it is some of the most fun tasks I have in my role. I like engaging people to hear their thoughts and making it a community event because I often get people that write back to me or comment. It is cool to have that thought partnership with a broader audience.

On Challenges With Product-Led Growth

Product-led growth is not easy. Each type of sales motion can be hard. Traditional top-down sales is very consultative and high-touch with buyers, which can be very long. Product-led growth can be hard in a different way. While targeting the end user, you need to build a product that very quickly shows value and solves a pain point. It would help if you also had an excellent understanding of how and why an individual uses it.

1. An interesting learning I have seen with PLG is the role of product analytics - businesses like Amplitude providing insights into product usage.

2. Another example is using that product data to make informed engagement with potential customers based on where they are in their product lifecycle or triggering events that would make them a good candidate for a purchasing decision. This category is often called PLG CRM, with businesses like Endgame operating in it.

3. The last thing which is hard about PLG is that it is often a community-led sale. I do a lot of open-source investing, which requires you to build a community around your open-source project, include them in the creation of the open-source, and listen to them as part of usually a steering committee. When doing PLG, there is a lot of engagement with the individual, and you need to make them an advocate or a champion of your work because people often make community-led decisions. If I know someone who has referred me to this product and said good things, my likelihood of buying is higher.

Overall, the grassroots approach to PLG can be very complicated.

On Community As The New Pre-Sales

The community has become the new pre-sales because people in the community are speaking on the vendor's behalf or negating the comments of the vendor. So if I am a prospective user who has not signed up and going to a community forum/Slack or Discord channel, I am going to read how other people perceive the product experience. Those community members are essentially acting as a pre-sales team for vendors, even though they do not intend to. The focus on the individual user or community member is crucial for businesses and requires you to cast a wider net of engagement. That is why you see the role of community managers and DevRel professionals coming to fruition over the past few years because they act as the champion of the user and build trust with them. It is not always about pitching the product. It is about situating the product within the context of the user's workflow or environment.

To me, a community can be a competitive moat. We have seen that with businesses like Databricks and Confluence that have active and thriving open-source communities. We can also see this with closed-source companies like LaunchDarkly that have well-attended conferences. The more people who are actively committing or using the products and speaking on your behalf, the less likely prospects will consider other alternatives.

On Hiring DevRel Professionals

A great DevRel individual is someone with a technical bend to them. They could be an engineer, a technical writer, or even a customer engineer. They can relate and empathize with the end user in order to translate their needs into technology. Here are several recommendations I have had with founders:

1. Go on LinkedIn and find people that may not have DevRel in their titles today but exhibit those personas.

2. Find people in an open-source community who are deeply passionate about the technology. Have them become advocates for it as part of your business.

3. Identify DevRel newsletters (like DevRel Weekly) that have a community of DevRel individuals and job postings.

For me, the DevRel role is slightly different within traditional closed-source SaaS and open-source. In the open-source community stand, you need to encourage them to contribute to the project and demonstrate their conviction in the technology. That is fundamentally not what happens with commercial SaaS products. But most of the role is similar:

• being the voice of the user

• translating their needs into product and sales

• positioning as a thought leader within the domain

• encouraging their contributors if it is open-source

• managing their expectations of the product

Two hacks that I have for early-stage founders:

1. Set up a Slack or Discord with each of your design partners and enable a central thread so they can ask general questions. This helps because when you make the forum public, you already have content in there. Prospective users will not land on something that's completely vacant.

2. Get on a weekly content cadence. Start to build a brand and reputation as a thought leader. That can also bring people into your community because they will go to you for advice not just about your technology but the general space you are operating in.

On Angel Investing

1. Tell people you are interested in angel investing. If you are at a late-stage company or an early-stage startup, tell your friends, tell your peers, tell the people on your board: "Hey, I am really excited about angel investing. Here are the three reasons why I could be additive to the team (community development, building and scaling up an engineering team, creating a cohesive team culture, etc.) So start telling people you are interested.

2. Build a personal identity or brand as an angel investor, leveraging your superpower. You can do that through social media or content marketing by sharing advice around particular topics or providing insights into technical shifts that are emerging.

3. Go through Crunchbase and look up companies that you admire and wish you were an investor in. See which other investors or angels are part of the cap table. Reach out to them directly to build a relationship. Something as simple as: "Hey, I really admire the work that you have been doing with this business. I think business X is very cool for these reasons. I would love to learn more about what got you excited and get any advice you have about me becoming an angel or potentially me working with you on angel opportunities."

It is pretty fun being an angel investor, so I encourage others who are interested to think through the three steps above.

On Being The Founding Partner at Quiet Capital

It was awesome being part of the Redpoint team. Working with Satish, Scott, Tomasz, and Alex was fantastic. For me, I got that little bug that founders do. I was engaging with all these founders who had a big vision, were trying to change the world, and were going from 0 to 1. I wanted to have more of that operator experience and empathy with the founders.

I got excited about Quiet Capital because it allowed me to fuse my learnings from Cisco/Redpoint and my own perspective of what modern venture looks like. I can build and scale up a VC fund, while contributing to the firm's tools, processes, hiring, and decision-making. It has been really awesome. We are all vertical specialists here. I lead the early-stage enterprise practice, while others focus on crypto, healthcare, fintech, etc. It is exciting to leave a fingerprint on an organization and a bit of legacy in Silicon Valley.

On Evaluating Early-Stage Entrepreneurs

Over my investing journey from public markets, to early-growth, to early-stage, I have become ever more focused on people, their technical aptitude, their market fit, their unique insight, their grit, and their make-anything-happen attitude. I spend much more time getting to know the founders, figuring out what motivates and makes them tick, and trying to dream with them about what could evolve from their business and what they want to be known for.

My mental checklist of the three buckets is: people first, technology second, and market third. It is important to have a good understanding of what technology they are building, what the differentiation is, what it will enable a user or a business, what the ROI of adopting this tech is, and how it fits into the current technical architecture or enables a re-platforming. But incredible people find a way. I am more driven than ever to meet with and get to know early-stage enterprise founders that are trying to disrupt their category or build a new space because they see something that other people do not.

On Modern Venture

For me, modern venture is a reflection of flexibility. The role of traditional Series A funds has evolved. We see them going earlier to Seed and pre-seed. We see traditional growth funds becoming early growth or going even as early as Series A. We see crossover funds going all the way down to Seed.

At Quiet, we prefer to lead and are open to being complimentary capital and getting to know founders deeply in order to add value, build trust, and continue to support the founders through their entire journey. We like to build a relationship and bring to bear $50-100M of capital for those next stages of growth. The flexibility to do that is exciting. Most Series A funds will only lead the round and take pro rata, but there is not this continuous support of the founder.

On Data-Centric ML

One trend I am super excited about is the idea of Data-Centric Machine Learning. Ten years ago, there were only Ph.D.s doing Bayesian modeling for ML. Since then, there has been an evolution of democratization of ML to data scientists and even analytics engineers. With that evolution, you have more open-source models that someone could take down, retrain, rebalance the weights, and deploy into production. Thus, algorithm development has become less challenging for an average use case. Of course, there has been crazy stuff going on in NLP with GPT-3 or Cohere, but for everyday data science where your team is trying to deploy a recommendation model, there are available resources to you.

That leads to the emphasis on the data layer, which people call Data-Centric ML. Tasks such as data capturing, data labeling, data balancing to understand which data point hurts/helps/has no effect on the model training, edge case detection to trigger retraining, or drift detection have become incredibly important to the ML development lifecycle. I am very excited about technologies like Unbox.ai that helps with ML data debugging and data quality to enable teams to use data more effectively for ML use cases.

When I think about the BI analytics stack, great data quality tools, such as Monte Carlo, look at the data warehouse to detect null values or distribution changes. This is very data-centric. The difference with ML data quality solutions is that they actually look at the training data, which may not always be in a cloud data warehouse. They tie to the trained models and frameworks. They can perform embeddings to understand clusters of datasets and highlight edge cases or mislabeled data. What is also very different is that they can tie into the model in production to understand edge cases or data drift. Overall, the point of integration, the workflow, and the end user are very different between Monte Carlo-style analytical data quality and ML data quality solutions.

Show Notes

• (01:56) Astasia shared her childhood growing up in Silicon Valley.

• (05:12) Astasia reflected on her undergraduate education at Stanford - studying Political Science and International Relations.

• (06:35) Astasia discussed her research at the Graduate Business School with Professor Condoleezza Rice on a case study called "San Leon Energy: Hydraulic Fracturing in Poland" - which explores how to manage the political risks of using a controversial energy extraction technology in the European Union.

• (09:26) Astasia talked about her year in the UK getting a Master's in Technology Policy at the University of Cambridge's Judge Business School.

• (12:52) Astasia recalled her experience as an Equity Research Analyst at Baird and Co.

• (17:49) Astasia mentioned her work at Cisco Investments, driving their cloud-infrastructure M&A and venture investments.

• (20:58) Astasia shared her thoughts on different M&A frameworks she learned from Cisco.

• (23:27) Astasia reflected on her decision to join Redpoint Ventures in early 2017, leading investments across developer tools, cloud infrastructure, data/ML infrastructure, AI applications, and cybersecurity.

• (25:44) Astasia debunked misconceptions about the venture industry.

• (29:30) Astasia discussed ways to prove her value upfront in potential deals and start forming her investment theses as a new investor.

• (33:01) Astasia dissected the key factors that triggered her to invest in the Series A of Solo.io and the Series B of LaunchDarkly (in the domain of cloud infrastructure).

• (38:48) Astasia explained her Series A investment in Hex and Series B investment in Preset (in the domain of data infrastructure).

• (44:12) Astasia shared advice she had given her portfolio companies in hiring decisions, pricing products, and navigating go-to-market strategy while at Redpoint.

• (47:36) Astasia walked through her process of writing comprehensive research primers in her Medium blog Memory Leak on wide-ranging topics - from data science notebooks and data orchestration to data pipelining and ML data management.

• (51:19) Astasia shared the typical challenges she has seen in companies looking to incorporate Product-Led Growth into their go-to-market motion.

• (54:10) Astasia discussed building a community as a fuel for product-led growth and shared advice to startups thinking about starting their community initiatives.

• (56:40) Astasia shared advice for hiring good DevRel practitioners.

• (01:00:15) Astasia shared advice for a smart, driven operator who wants to explore angel investing.

• (01:03:26) Astasia talked about her current journey as the Founding Partner at Quiet Capital, sitting on its early-stage enterprise team and leading opportunities across pre-seed, seed, Series A, and Series B.

• (01:05:13) Astasia expanded upon her typical mental checklist to evaluate entrepreneurs and make investment decisions.

• (01:07:36) Astasia briefly touched on LP fundraising for Quiet Capital to become a "modern venture firm."

• (01:09:59) Astasia emphasized her enthusiasm for the Data-Centric ML movement.

• (01:13:41) Closing segment.

Astasia's Contact Info

• LinkedIn

• Medium

• Twitter

Quiet Capital

• Website

• LinkedIn

• Twitter

Mentioned Resources

Content

• John Gannon Blog

People

• Satish Dharmaraj (Redpoint Ventures)

• Scott Raney (Redpoint Ventures)

• Amanda Robson (Cowboy Ventures)

Notes

My conversation with Astasia was recorded back in April 2022. Since then, many things have happened. I'd recommend:

1. Signing up for her Memory Leak newsletter

2. Browsing through Quiet Capital's new portfolio careers page

3. Listening to Astasia's appearance on the Data Stack Show

4. Checking out Quiet Capital's investments in Edge Delta, Diagrid, and Omni

5. Looking at her real-time infrastructure landscape

About the show

Datacast features long-form, in-depth conversations with practitioners and researchers in the data community to walk through their professional journeys and unpack the lessons learned along the way. I invite guests coming from a wide range of career paths — from scientists and analysts to founders and investors — to analyze the case for using data in the real world and extract their mental models (“the WHY and the HOW”) behind their pursuits. Hopefully, these conversations can serve as valuable tools for early-stage data professionals as they navigate their own careers in the exciting data universe.

Datacast is produced and edited by James Le. For inquiries about sponsoring the podcast, email khanhle.1013@gmail.com.

Subscribe by searching for Datacast wherever you get podcasts, or click one of the links below:

• Listen on Spotify

• Listen on Apple Podcasts

• Listen on Google Podcasts

If you’re new, see the podcast homepage for the most recent episodes to listen to, or browse the full guest list.