AI-Driven Auto Shops with Mastertech.ai and Data Observability with Monte Carlo | E2035

30 Oct 2024 (2 months ago)
AI-Driven Auto Shops with Mastertech.ai and Data Observability with Monte Carlo | E2035

Alex Wilhelm kicks off the show (0s)

  • A query is attempted to diagnose an issue with a 2019 Subaru Outback's passenger window that gets stuck halfway up (6s).
  • Two technical service bulletins (TSBs) were found to be related to power window issues for this model, providing a starting point for diagnosing the problem (15s).
  • The podcast is sponsored by Vanta, which offers compliance and security solutions for startups, with a limited-time offer of $1,000 off for listeners (28s).
  • Squarespace is also a sponsor, offering a free trial and a 10% discount on the first purchase of a website or domain with the code "twist" (46s).
  • Ki is another sponsor, allowing users to trade on US elections and offering $20 when depositing $100 (1m0s).
  • The host, Alex, introduces himself and welcomes listeners back to "This Week in Startups" (1m12s).
  • Alex introduces the guest, Linda Gray, a founder with deep technology experience who has applied her skills to a different industry through her company, Mastertech.ai (1m24s).
  • Linda Gray expresses her gratitude for the support from Launch and her excitement to be on the program (1m40s).

Linda Gray's career and Mastertech.ai origin (2m5s)

  • Linda Gray spent 15 years at Microsoft, holding various roles including principal software engineering lead manager, before working at Niantic, the company behind Pokémon Go (2m5s).
  • Gray's career path was not a straight line, and she made a pivot from Microsoft and Niantic to Mastertech AI, which focuses on building AI for mechanics and frontline workers in auto shops (2m32s).
  • Gray joined Microsoft out of college, interning in 2004, and rose through the ranks to become a principal engineer on the Outlook web team and later led teams at Microsoft Outlook, Microsoft Teams, and Xbox Live (2m53s).
  • After 15 years at Microsoft and 17 years in tech, Gray wanted to do something new and make a real-world impact, which led her to consider starting her own company (3m35s).
  • Gray was motivated to start a startup and work on a zero-to-one project, taking advantage of the innovation happening in AI and its potential to make a real-world impact (4m30s).
  • The decision to apply AI to auto shops and mechanics was a deliberate choice, but the specific reasons behind this choice are not mentioned in this part of the conversation (5m6s).

Mastertech.ai and the auto repair industry (5m17s)

  • The auto repair industry was chosen as the focus for a startup due to its underserved and overlooked nature, presenting an opportunity to make a significant real-world impact by bringing technology to this market (6m18s).
  • The idea of working with Frontline workers, such as those in auto repair shops, was influenced by Microsoft's efforts to bring products to Frontline workers, which was observed during the founder's time at the company (7m0s).
  • The founder's experience on the Microsoft D Tod platforms team, specifically working on the apps ecosystem for Teams, provided opportunities to integrate with third-party apps and indirectly serve Frontline workers (7m8s).
  • The approach to problem-solving in big tech companies, such as Microsoft, often involves building horizontal platforms that cater to a wide range of users, resulting in a "least common denominator" experience (7m31s).
  • In contrast, the founder wanted to create a solution that was tailored to a specific problem, in this case, the auto repair industry, and utilize technology to its fullest potential (7m55s).
  • The decision to start a company focused on the auto repair industry was motivated by a desire to move away from the "tech bubble" and solve problems that have a significant impact on the real world (5m50s).
  • The founder's goal was to create a solution that would have a bigger real-world impact by bringing technology to underserved markets, rather than solving problems for other tech companies and engineers (6m29s).

Vanta - Get $1000 off your SOC 2 (8m6s)

  • To sell to bigger customers and increase average contract value (ACV), it is necessary to clear compliance checks, including SOC 2, which ensures companies keep their customer data safe (8m8s).
  • SOC 2 is a standard that companies must meet to operate at the highest end of the market and land big deals, including lighthouse customers (8m23s).
  • Without SOC 2 compliance, companies may not be able to land big deals and operate at the highest end of the market (8m32s).
  • Vanta makes it easy for companies to get and renew their SOC 2 compliance, with customers becoming compliant in just 2 to 4 weeks on average (8m39s).
  • Vanta automates compliance for GDPR, HIPPA, and more, saving companies hundreds of hours of work and up to 85% on compliance costs (8m48s).
  • Using Vanta can help stop slowing down sales teams and allow companies to sell to bigger customers (9m1s).
  • Vanta is offering $1,000 off SOC 2 compliance at v.com/twist, and using the code can result in $11,000 off (9m3s).

Technological adoption and Mastertech.ai's benefits (9m11s)

  • The average car shop's level of technological savviness varies, with shop owners being more resistant to change, while technicians are generally tech-savvy and open to adopting new technologies (9m51s).
  • Many technicians are from Gen Z and are familiar with the latest technologies, but there hasn't been a great tool or application built for them until now (10m14s).
  • Mastertech.ai was co-founded by Dave, a 20-year veteran of the auto repair industry, and his partner, who met to discuss the application of AI in the industry (10m24s).
  • Mastertech.ai aims to provide a tool that works on the ground in auto shops, using AI to help technicians navigate digital information and technical specifications for various vehicles (10m50s).
  • The adoption of Mastertech.ai has been successful, with engagement increasing week over week, and it is being used by technicians in shops (10m58s).
  • Mastertech.ai is a software that uses AI to provide technicians with the information they need to diagnose and repair vehicles, regardless of the make, model, or year (11m30s).
  • The job of technicians and mechanics involves navigating digital information and technical specifications, as well as hands-on work, with an average of 25% of their time spent on computer research (11m45s).
  • Mastertech.ai aims to reduce the time spent on research by providing a single platform for technicians to access the information they need, increasing precision and reducing the risk of errors (12m21s).
  • The platform is analogous to healthcare, where precise information is critical to making accurate diagnoses and treatments (12m40s).
  • AI investments in healthcare are being used to coalesce patient history and diagnostics for doctors, and a similar approach is being applied to vehicles, utilizing the unique blueprint of specifications that comes with each vehicle (12m46s).
  • The data for vehicles is gathered from licensed OEM data providers, who license out the data on behalf of the OEMs, and approval for the use case is obtained from each OEM, a process that has been ongoing for the past year (14m30s).
  • The majority of the necessary approvals have been obtained, and the data is directly from the OEMs for procedures, specifications, and other information (14m49s).
  • AI is used to navigate the user's intent, serve the correct data, and assist technicians in the most effective way possible (14m58s).
  • The AI component allows technicians to ask questions, which are then parsed and turned into a query to retrieve the relevant information from the OEM database (15m10s).
  • The system uses semantic search, allowing for more flexible and user-friendly searches, rather than strict keyword searches or file folder lookups (15m28s).
  • Every answer provided by the system is backed up by the original manufacturer's source, ensuring accuracy and trustworthiness (15m54s).
  • While it is not necessary to have agreements with all OEMs, having a critical mass of around 80% of OEMs is considered sufficient to provide a comprehensive service (16m9s).

Mastertech.ai's OEM approval process and community data (16m18s)

  • Mastertech.ai is working on getting OEM approval from various car manufacturers, with the goal of providing valuable data to average auto shops, and they currently have approval from most major manufacturers except for Honda and Toyota (16m18s).
  • Honda and Toyota are significant manufacturers, but they are not as big of a priority for Mastertech.ai's primary customers, who are repair shops that often specialize in European vehicles (16m40s).
  • Subaru has already come to terms with Mastertech.ai, indicating that the issue is not with Japanese car companies in general, but rather with getting individual approvals from each OEM (17m23s).
  • Mastertech.ai is focusing on three major sets of data: OEM data, community data, and user-submitted content, with the goal of creating a platform that combines OEM information with real-world experience and knowledge from technicians (17m55s).
  • Community data is critical in the automotive repair industry, as technicians rely heavily on personal experience and human knowledge to diagnose and fix problems, and Mastertech.ai plans to launch a user-submission content pipeline to collect and share this type of data (18m12s).
  • The vision for Mastertech.ai is to become the "Stack Overflow for automotive repair," a platform where technicians can share knowledge, leave notes, and provide breadcrumbs for others to follow, making the job faster, easier, and safer (19m10s).
  • The platform aims to accumulate human intelligence and experience from technicians, rather than relying solely on artificial intelligence, to help make the job of automotive repair safer and more efficient (20m34s).
  • Auto technicians face a high-pressure and dangerous job, with limited software help and time to find necessary information, and Mastertech.ai's platform aims to address these challenges by providing a comprehensive resource for technicians (20m3s).

Squarespace - Use offer code TWIST to save 10% off your first purchase of a website or domain (20m50s)

  • Squarespace is a platform where users can build beautiful websites with amazing design and functionality without needing to hire developers or a web shop, which can be expensive and unreliable (20m50s).
  • Building a website with a web shop can cost thousands to tens of thousands of dollars, and there's a risk of the developers disappearing or ghosting the client (21m8s).
  • Squarespace is a trusted platform that has been used for over a decade to maintain and evolve web presence, and it's suitable for various businesses, including startups, schools, banks, projects, and books (21m17s).
  • Squarespace has a design intelligence tool that uses AI to build custom websites that are gorgeous and easy to create (21m34s).
  • In addition to design, Squarespace also offers built-in payment technology, analytics, and domain registration to support businesses in selling products and tracking sales (21m48s).
  • Users can sign up for a free trial on Squarespace and get 10% off their first website or domain purchase by using the offer code "TWIST" at squarespace.com/twist (22m4s).
  • Squarespace is a long-running partner of "This Week in Startups" and is appreciated for its generosity and support (22m15s).

Mastertech.ai's AI-driven diagnostics demo (22m20s)

  • The MasterTech AI platform is a web application that is fully mobile optimized and allows users to add vehicles they are working on, with plans to build a fully native mobile application in the future (22m43s).
  • The platform features a chat-first interface where users can get assistance with various tasks, including diagnosing issues, looking at specifications, and managing labor times (23m25s).
  • Users can scan in vehicles using a VIN scanner or camera-based VIN scanner, and the platform provides quick actions for common tasks (23m9s).
  • The platform uses AI to navigate OEM procedures, such as noise diagnosis, and provides detailed flowcharts and known issues reported by the National Highway Transport Safety Administration (24m2s).
  • The system can detect generic issues and prompt users for more information to narrow down the problem, allowing users to provide brief descriptions of the issue (24m52s).
  • By adding more specific information, the platform can provide more likely causes of the issue and known issues reported for the specific engine or model (25m27s).
  • The platform aims to help technicians by providing a centralized source of information, eliminating the need to search multiple database sources (25m56s).
  • A query is performed to find information on a 2019 Subaru Outback with a V6 engine, specifically why the passenger window gets stuck halfway up, and the system returns results showing known issues with a faulty power windows switch and a mechanical issue with the window regulator, as well as related technical service bulletins (TSBs) and a Subaru procedure for resetting the module (26m19s).
  • The system's ability to quickly and accurately provide comprehensive information is highlighted as an example of how technology can improve productivity in the real world (26m2s).
  • The company, Mastertech.ai, has seen significant traction in the market since launching publicly to shops on May 1st, with around 40 shops signed up on a monthly subscription and increasing user engagement over time (27m55s).
  • The company's AI-based platform allows them to know exactly what users are looking for and whether they were able to help, which helps prioritize data needs and improvements (28m54s).
  • The company is generating around $100,000 in annual recurring revenue (ARR) with an average monthly subscription price of $180 per shop, and is seeing accelerating momentum as they incorporate more data and achieve better product-market fit (29m17s).
  • Word of mouth is becoming a significant source of new customers, with shop owners recommending the product to others in online groups (29m41s).
  • MasterTech.ai is working on providing auto shops with the necessary data on the ground, including procedures, specifications, fluids, DTC codes, diagnosis help, labor times, wiring diagrams, maintenance schedules, and shop management data to help with recordkeeping and customer history navigation (30m1s).
  • The company is growing its product and seeing better engagement and attraction, with plans to participate in coaching groups for shop owners and attend trade shows to aid growth (30m31s).
  • The rise of electric vehicles (EVs) and hybrids is expected to add complications for auto shops, but MasterTech.ai sees potential in helping shops transition to servicing EVs and hybrids with its platform (31m16s).
  • EVs eliminate some maintenance associated with traditional internal combustion engine vehicles, but they also introduce new complications that most shops are not equipped to handle (31m19s).
  • MasterTech.ai aims to leverage its platform to help shops with the transition to servicing EVs and hybrids, and is looking to access service data and onboard diagnostics remotely (32m1s).
  • The company is envisioning a future where its platform can be translated to other verticals, such as HVAC, where service information is also a challenge (33m10s).
  • MasterTech.ai can be found online at mastertech.ai, and the company is excited about its future prospects and potential applications in various industries (32m57s).
  • A potential application of AI-driven auto shops is the ability to scan the serial number of machinery, such as HVAC units, cars, or boats, to access necessary service information, including blueprints and wiring diagrams, to aid technicians on the ground, utilizing AR and voice technology (33m39s).
  • This concept represents a future vision for the industry, with potential applications in various fields that require extensive maintenance and have high value, such as solar panel installation, wind power, and batteries for grids and homes (34m3s).
  • The idea of AI-driven auto shops, or "Master Tech," has the potential to expand into different verticals in the future, encompassing various industries that require significant maintenance and have high value (34m21s).
  • The discussion highlights the vast possibilities and potential applications of AI-driven auto shops, with the need for further exploration and development of this concept (34m4s).
  • The conversation concludes with an appreciation for the guest's time and a recognition of the significance of the idea presented, with a call to action to further develop this concept (34m27s).

Kalshi—the largest regulated predictions market—now lets you trade on US elections. Visit to see live odds, trade, and get $20 when you deposit $100. (34m34s)

  • Prediction markets allow people to buy and sell contracts on future events with prices set by market opinion, making them a brilliant way to gauge the likelihood of outcomes (34m35s).
  • Kalshi, the world's largest regulated predictions market, has made it legal to trade on the upcoming US elections, allowing US citizens to place trades on election outcomes for the first time in 100 years (34m49s).
  • Unlike traditional political polls, prediction markets can provide more accurate insights into the likelihood of election outcomes by showing where "sharps" (smart people) are putting their money (35m6s).
  • Election outcomes have significant impacts on taxes, the economy, and the world of investing, making it crucial for individuals to be able to hedge their bets against uncertainties (35m11s).
  • Traditionally, only large corporations and the wealthy could hedge their bets against election uncertainties, but Kalshi now allows anyone to protect themselves if their preferred candidate loses (35m19s).
  • Kalshi provides real-time election odds, allowing users to check the current market opinion and make informed decisions about trading (35m25s).
  • New users can sign up for Kalshi at ki.com/twist and receive an additional $20 with their first deposit of $100 (35m40s).

Data observability and AI's role in anomaly detection (35m51s)

  • Monte Carlo is a key company in the data observability movement, a term they coined a few years ago by applying the concept of observability from devops and software engineering to the data and analytics space (36m45s).
  • The company was founded to help data engineers and teams manage reliability and quality in their data systems, as they previously lacked the necessary tooling and methodologies (37m19s).
  • Monte Carlo's goal is to provide data engineers with the equivalent of tools like Datadog or New Relic, but for data systems, to ensure high-quality products and reliable delivery to end-users (38m1s).
  • The company has grown significantly, now serving over 400 enterprises across various industries, including tech, pharma, finance, manufacturing, sports, and education (38m46s).
  • Despite increased competition in the data observability space, Monte Carlo views it as a positive development, as it signals to customers that the category is important and drives innovation (39m31s).
  • The company's co-founder and CTO, Leor Gish, believes that competition makes them better and that they don't have a monopoly on good ideas, which ultimately benefits their customers (39m25s).
  • There is real interest in data observability, and it is a net positive, with many companies evaluating solutions and competing in the space (39m33s).
  • The concept of data observability has broken out of the tech space and is now well known in the broader world of business, including industries such as pharmaceuticals (40m18s).
  • The term "data downtime" was key in resonating with people and explaining the problem that data observability solves, which is serving the wrong data to end users (40m51s).
  • Data downtime is the problem, and data observability is one solution for that problem, which involves educating people about the issue and the solution (41m29s).
  • Gartner has picked up the terminology, and market guides are now using it, indicating that data observability has become a widely recognized concept (41m46s).
  • Many enterprises are now putting out RFPs for data observability, showing that they understand the concept and want a solution (42m6s).
  • Monte Carlo keeps tabs on data flowing through a company's pipes and can spot anomalies, such as a data point coming in at zero when it's never been zero before (42m25s).
  • Machine learning (ML) and artificial intelligence (AI) are applied to the process of seeing anomalies and other issues as they happen, with a significant amount of ML being used (42m40s).
  • One key innovation in the space is the idea that data engineers and analysts can't be expected to track every single table and field in their databases and data warehouses, and that a solution is needed to help them understand how the data behaves and what it means for it to be broken (42m58s).
  • Machine learning (ML) and artificial intelligence (AI) are being used to scale data monitoring, allowing a small group of people to handle a large amount of data by analyzing patterns, predicting what the data should be, and alerting when it breaks from that pattern (43m18s).
  • AI techniques can also be used to analyze metadata around the data, such as descriptions and logs, to help inform how to spot breakages and issues (43m48s).
  • AI is used to help people get to the root causes of problems quicker, finding out what happened and why, and where the problem is originating from in complex data pipelines (44m8s).
  • The applications of ML and AI are numerous, and the company has been using AI to power the engine of its product (44m25s).
  • Despite being well-positioned to capitalize on the current AI boom, the company has chosen not to raise large amounts of funding, having already been well-capitalized through previous rounds (44m46s).
  • The company, Monte Carlo, is excited about its ability to help customers build AI, in addition to using AI within its own product (45m10s).

AI's accelerant effect on Monte Carlo's growth and strategy (45m15s)

  • The current AI models are considered incredible and a commodity, with almost every company having access to them through API keys from companies like Open AI, Anthropic, and others (45m17s).
  • The real differentiator for companies is the data they can inject into these models, and that's where Monte Carlo fits in by helping companies build pipelines that power AI applications and analyze structured data (45m40s).
  • These pipelines can break, and Monte Carlo's customers use the platform to monitor, alert, and prevent downtime in those pipelines (46m2s).
  • The concept of data downtime is critical, especially when AI models ingest data to interact with customers, and any errors can have significant consequences (46m14s).
  • Monte Carlo is an AI and data observability company that helps customers build more AI-driven applications and ensures they have the right data to support those applications (46m35s).
  • The growth of AI has been a significant accelerant for Monte Carlo, with close to 100% of their customers asking about AI use cases when considering solutions (47m2s).
  • Many customers are either building AI applications or planning to invest in them and want to ensure their data durability provider can support those efforts (47m10s).
  • Monte Carlo has seen an increase in requests to monitor unstructured data, which has become more accessible to analyze with the commoditization of AI models (47m20s).
  • The company has seen cool use cases with unstructured data and is working with customers to ensure their pipelines are working effectively (47m49s).
  • The growth of generative AI has been a booster to Monte Carlo's business outcomes, although the exact impact on their growth rate is unclear (48m1s).
  • Monte Carlo works with various data platforms, including data lakes, data warehouses, and data bricks, but has not been acquired by any of these companies, including data bricks (48m12s).
  • Monte Carlo has a good partnership with data bricks and has never considered selling the company (48m32s).
  • The company Monte Carlo had raised a Series D round of $135 million with a $1.6 billion valuation in May 2022, with plans to invest in engineering, data, product, and go-to-market work, but the world's economic situation changed shortly after, advising companies to reduce spending and growth rates (48m56s).
  • Despite the change in the economic situation, Monte Carlo's plans remained largely unchanged, as they had always intended to build a long-lasting business and spend money responsibly, considering the business's needs and unit economics (50m1s).
  • The company's philosophy is to spend responsibly and prioritize customer happiness while keeping costs within reason, which has helped them navigate economic cycles and make adjustments as needed (50m52s).
  • Over time, Monte Carlo has made mistakes in hiring and resource allocation, but they have learned from these experiences and continued to grow (51m8s).
  • The company has seen economies of scale, particularly in terms of gross margin and unit economics, which have improved over time due to a combination of natural scaling and active efforts to optimize infrastructure spend and code (51m40s).
  • Monte Carlo's goal is to continue improving their gross margin every year, and they have been successful in doing so, with a positive trend as they scale (51m54s).
  • Monte Carlo recently hired its first Chief Revenue Officer, Tim, who has experience in scaling playbooks across larger teams, most recently at Stack Overflow, where he helped make Stack Overflow's alternative AI business. (52m17s)
  • The hiring of a Chief Revenue Officer is seen as a scaling play for Monte Carlo, as the company needs to get more consistency and execute the playbook it has learned across the board. (52m32s)
  • The role of the Chief Revenue Officer is not just a revenue-generating focus, but also involves bringing in a responsible adult to take the company in the right direction and ensure consistency and execution across the team. (52m35s)
  • Monte Carlo has been added to the Twist 500 list, which features companies that are expected to have the biggest financial outcomes, a proxy for innovation, market disruption, and more. (53m35s)
  • Despite having access to lots of capital, Monte Carlo's founder acknowledges that failure is still a possibility, not in the sense of going out of business, but in failing to meet the company's grand ambitions, such as building an independent business that could eventually go public. (54m13s)
  • The founder believes that the odds are against startups succeeding, but Monte Carlo is growing and has a real pain and need in the market that it is serving, making it unlikely to go away anytime soon. (54m49s)
  • Monte Carlo is seen as a company that started its "picks and shovels" business before the current AI gold rush, and is well-positioned to take advantage of the growing demand for AI solutions. (55m22s)
  • Leor from Monte Carlo was a guest on the show, and the host expressed appreciation for his appearance (55m27s).
  • The host invited Leor to return to the show next year to provide an update on his progress (55m33s).
  • Leor mentioned the website Monte Carlo Data (dat.com) for those interested in learning more (55m35s).
  • The host, Al, and Leor exchanged gratitude and farewells at the end of the conversation (55m39s).

Overwhelmed by Endless Content?