The video provides advice and insights from successful founders on various aspects of building and growing a startup, including prioritizing speed and iteration, balancing security and scalability, effective communication and management techniques, and hiring strategies.

• 🚀
  00:00

  Jared Frieden and his panel of successful founders give advice to startup founders, Diana is building a back-end technology for augmented reality, and Second Measure provides insights into company performance through credit card transactions.

  • Jared Frieden and his panel of successful tactical founders discuss technology and give advice to startup founders, with questions from the Startup School forum.
  • Plain Grid Clean Grid is a software company that provides collaboration and version control tools for the construction industry, similar to Github.
  • Segment is a single API that collects and organizes customer data into hundreds of downstream tools, built entirely on AWS infrastructure and with a tech stack that includes Python, Go, Java, Kotlin, Objective-C, and React.
  • Diana is the founder and CTO of SEO Reality, which got acquired by Niantic, and they are building a back-end technology for augmented reality to enable developers to build easy AR experiences with advanced features.
  • Second Measure analyzes billions of credit card transactions daily to provide insights into public and private company performance for clients including VC firms, hedge funds, and big brands.
  • The speaker's company has a technical team of 50 people, split evenly between data scientists and engineers, and their core product is data, which requires them to rapidly query and run complex queries on their backend using AWS services like Lambda and Spark, and a React-based front-end with columnar data stores to service queries.
• 💡
  08:17

  Building a blueprint viewer for iPad in 2011 was challenging, but it inspired the speaker to create the first mobile viewer using their graphics experience.

  • Start-up school attendees are asked to share the story of their v1, including who built it, how it was built, how long it took, and what went wrong in the process.
  • Putting blueprints on an iPad in 2011 was challenging due to the low computer availability and lack of graphics chip, causing crashes and memory overruns.
  • The first prototype was slow, but the challenge inspired the speaker to use their graphics experience to create the first blueprint viewer for mobile.
  • The first prototype lacked a web interface and had a delete button next to the publish button without confirmation, so the team managed everything on the back end for the first 30-40 customers.
  • The founders initially built a classroom lecture tool to help professors get feedback, but it failed, leading them to switch to building a competitor to Mixpanel called Segments.
• 💡
  12:55

  After struggling to find success, the founders turned their internal tool into a successful product, while the speaker and their team utilized slam algorithms to create an augmented reality experience for gaming, which eventually led to acquisition by Niantic.

  • After burning half a million dollars and being at square one, the founders turned their internal tool into a product which became successful.
  • They cleaned up the library, launched it on Hacker News, got a thousand stars, built a v1 in a week, and got the seed of product market fit and a user base.
  • The speaker and their co-founder utilized slam algorithms, commonly used in robotics, to create an augmented reality experience on phones due to the advancements in phone computation power.
  • The initial version of the AR experience was discouraging due to low frame rate, but after removing research code, it started working at 30 frames per second, a year and a half before Air Kid launched.
  • The team interviewed potential users and found a market fit in gaming, and despite Apple's announcement of a similar product, they decided to double down and accelerate their roadmap to launch their product.
  • The speaker and their team developed a product using slam algorithms, which took about a year of part-time work and six months of full-time work, and after getting acquired by Niantic, they were able to accelerate the process with the help of two more engineers.
• 🚀
  19:51

  Prioritizing speed and iteration over best practices, a startup shifted towards implementing unit tests and formalized code reviews as they grew and gained revenue.

  • The service started by focusing on providing investors with predictions and data, but ultimately landed on building their own self-service platform with opinions and guidance for quick access to insights.
  • In the early days of a startup, it's more important to prioritize speed of development and constant iteration over engineering best practices like good test coverage and scalability.
  • The company implemented unit tests, CI/CD, and formalized best practices around code reviews and testing, while building more controls into their system to prevent breaking things, due to the growth and complexity of their systems and technical staff.
  • The speaker describes the three distinct phases of their development, starting with hackathon mode where they had no tests or infrastructure, followed by a private beta with duct tape code, and finally a complete rewrite with best practices after joining a larger company.
  • The company initially focused on rapid iteration and getting more users, but shifted towards improving development processes and implementing best practices as they brought on more engineers.
  • As the company gained revenue and customers, they realized the importance of securely handling data for their enterprise customers who rely on their services to power their business.
• 📈
  30:45

  Plain Grid Klinger focuses on security and scalability in their construction technology services, balancing between over-engineering and dealing with scaling issues to ensure their usefulness and prevent negative impacts on their customers' businesses.

  • Plain Grid Klinger's focus on security and scalability in their construction technology services to ensure their usefulness and prevent negative impacts on their customers' businesses.
  • Balancing between over-engineering and dealing with scaling issues, the service architects its system to last through the foreseeable future while trying to find a trade-off between engineering a year ahead of time and not three years ahead of time.
  • The speaker reflects on the technical debt of their engineering organization and acknowledges the challenge of avoiding spaghetti code, but notes that experienced developers generally learn to write cleaner code over time.
  • Functional testing is crucial for double-checking during integration tests, as unit tests can often become spaghetti code and waste time without actually testing anything.
  • Adopting engineering methodologies is important, but finding what works best for your team and constantly evolving your approach as your company grows is key.
  • The company has evolved from having zero documentation and no set process to implementing weekly planning and sub-teams to efficiently communicate and tackle different areas of focus with a larger team.
• 👨‍💻
  38:39

  💬 Effective communication, timeboxing, and lightweight management techniques are essential for engineering teams to stay on track and connected, while having a process and asking each team member individually about their predictions can help with project management and meeting deadlines.

  • The company allows individual engineering teams to self-organize and use their preferred tools, but implements a company-wide model of objectives and key results (OKRs) to measure progress and has weekly meetings to plan and discuss important problems.
  • Effective communication, timeboxing, and lightweight management techniques such as one-to-ones and stand-ups are essential for engineering teams to stay on track and connected.
  • Tips for working with non-technical co-founders include using them as testing opportunities and learning how to accurately estimate and communicate deadlines.
  • Having a process and asking each team member individually about their predictions for the next few weeks can help with project management and meeting deadlines, especially when dealing with non-technical co-founders.
  • The speaker's non-technical co-founder focused on external communications while the speaker, as the technical founder, owned the product and development, with the challenge being to balance engineering, product, and business trade-offs and communicate clearly with deadlines.
  • As a startup gains traction, the focus shifts from building a product for the market to building an organization and a system, and it's important to consider outsourcing only non-critical projects while investing in core competencies and IP.
• 💼
  49:52

  Hiring remote engineers can provide access to talented individuals, but local teams are still preferred due to the complexity of the technology, while hiring contractors can be beneficial but requires significant management time.

  • The company initially hired a local team but has recently started tapping into remote engineers for advantages such as accessing other talent pools and improving documentation, but still predominantly chooses to build the team locally due to the complexity of their technology.
  • Segments started by hiring full-time employees and later transitioned to hiring mostly remote workers, which came with access to talented individuals but required extra effort in communication and product alignment.
  • Hiring contractors can be beneficial for a company, as they can provide necessary skills and can even become valuable long-term employees.
  • Hiring contractors requires significant management time and remote work can be beneficial for a company.
  • Focus on getting product-market fit as soon as possible by finding your user base and getting in front of them all the time.
  • Get feedback from users early and often by showing them versions of your product, even if you plan to throw them away, to ensure you are solving their problem before going all in.
• 💻
  58:03

  Choose a cross-platform development library and start hiring developers earlier to build faster, and get in front of users as much as possible to receive powerful feedback.