Inside Ode with Anthropic, the startup betting AI services are the future of enterprise| Equity

Introduction to Ode with Anthropic and Its Mission

• Ode with Anthropic is a new venture launched by former co-founders of Fractional AI, Chris Taylor and Eddie Segal, which is backed by heavy hitters like Anthropic and Blackstone, and aims to help non-AI companies adopt AI technology the right way 10s.
• The company's mission is to close the gap between non-AI companies and the adoption of AI technology, which requires top-caliber applied AI talent that most companies do not have, and Ode is being founded to provide AI native services to these companies 4m42s.
• Ode is working with a mix of clients, including Blackstone portfolio companies, Hellman & Friedman portfolio companies, Goldman Sachs portfolio companies, and Anthropic clients, to help them innovate and drive business impact using AI 8m30s.
• From a technical perspective, Ode is building and delivering custom bespoke pieces of software that fit into the clients' problems and infrastructure, using the new AI technology to solve these problems, which involves challenging engineering, complex systems, and understanding of bespoke processes 12m6s.
• The company is aiming to build the defining company of the AI services market, with a close relationship with Anthropic, and is exploring engagements with various portfolio companies to use AI to innovate and drive outsized business impact 6m30s.

Ode's Client Base and Technical Approach

• A company that makes software in the governance, risk, and compliance space is working on a project to automate the customization and configuration process of their software, which currently takes months and delays the time it takes for customers to get value, with the help of AI services 10s.
• The AI system being built takes input from various sources such as calls with customers, PDF documents, and email exchanges, and helps the company to pre-plan and apply configuration internally, making the process much faster 42s.
• The implementation of software or configuration is a bottleneck for many businesses, including LogicGate, a software company that approached Anthropic to automate many tasks involved in implementing their software, which is a key bottleneck for their business 2m6s.

The Process of Implementing AI Solutions with Anthropic

• The process of working with Anthropic to implement AI solutions is very hands-on and changes throughout the life cycle of an engagement, starting with a company coming to them with hypotheses about where AI can help, but not necessarily having every detail in place 6m34s.
• The goal of Anthropic's work with companies like LogicGate is to automate many tasks involved in implementing software, creating leverage for their teams, and helping clients get more value out of the software, which can take several months to achieve 4m21s.
• The CEO of LogicGate reached out to Anthropic to explore the opportunity to automate many tasks involved in implementing their software, which is currently a key bottleneck for their business, delaying revenue and growth 3m15s.
• The process of working with clients involves solidifying the business case and understanding the potential impact of AI on their operations, including assessing the availability of data and the ability of teams to accomplish the desired tasks 10s.
• The implementation phase involves collaborating with the client's teams to build and refine the AI solution, with the goal of delivering something of value within a few months, and the long-term goal of creating a robust product feature that can be given to customers 2m6s.
• The timeline for achieving this goal can vary, but one rule of thumb is to have something in production that adds measurable business value within 3 to 6 months, to avoid the risk of the project becoming a perpetual endeavor with no return on investment 2m6s.

Measuring Success and Business Impact of AI Projects

• To measure the success of the AI solution, various metrics can be used, including revenue growth, efficiency, and time to value, with the goal of creating a business case that has a tangible impact on the business 4m42s.
• Different projects may have different metrics for success, but the key is to identify a clear opportunity for AI to drive business value and to establish a way to measure the results, which may involve shaping the measurement process to ensure that all stakeholders are aligned 6m10s.
• The ultimate goal is to create a solution that can drive revenue growth, improve efficiency, and increase the ability of the client's teams to handle more customers or engage in high-leverage work, with the potential to show up in the company's share price or revenue growth 8m20s.
• The primary goal of many businesses is to achieve revenue growth with some efficiency gains, which can be accomplished by removing the primary bottleneck of the business to unlock more revenue growth, ultimately resulting in a business that makes more money in terms of both top line and EBITDA 10s.
• To ensure the success of AI projects, it is essential to work closely with clients, combining AI expertise with their domain expertise, and to track a North Star for every project, checking in on it at least once a week to identify blockers and design solutions 2m6s.
• Measuring the success of AI projects is critical, and this can be done by building a simple system, creating a measurement system to compare output to known good output, and consistently measuring performance over time to identify areas for improvement 4m30s.
• Evaluations, or evals, play a crucial role in AI projects, as they help to identify edge cases, measure performance, and correlate it with business impact, such as revenue or time saved, allowing for data-driven decision making 6m20s.

Leadership Support and Project Prioritization

• The success of AI projects also depends on senior-level buy-in from the company, as these projects often require significant changes to business operations, and having the support of the CEO and other top executives is essential for implementation and adoption 9m40s.
• Many of the AI projects being worked on are top priority for the CEO and the company, involving critical product features or business processes, making the work exciting and high-stakes 11m20s.
• The team at Ode gets to work on the most exciting projects with senior-level support and deep subject matter experts, allowing them to be choosy about which projects they take on, with a primary focus on the size of the impact they can have, 10s.

Ode's Project Selection and Team Composition

• Ode applies a lens to determine which projects to take on, considering the top priorities of a CEO of a big company where AI can provide huge impact, and they require a team with the best applied AI engineers, product managers, and experience in doing these projects, 42s.
• The improvement of frontier models is a rising tide that helps with enterprise AI success, but model selection is just one ingredient in a system that has to be engineered, and the majority of time is spent on other aspects, 2m6s.
• When it comes to model selection, Ode is owned by Anthropic and will always choose Claude first if it works well enough, but they are model-agnostic and will consider other options if necessary, 10s.

Model Selection and Anthropic's Tools

• Anthropic's Claude Tech is an application that allows Claude's models to live inside Slack and run workflows independently, acting as a coworker, and Ode has had a positive experience with it, using it to run internal business processes, 2m6s.
• The collaboration between Ode and Anthropic, as well as between the application layer and model layer, yields a good outcome for the business, with Claude Tech taking on primary responsibility for operational requests and delegating to humans when necessary, 2m6s.

Market Demand and the Rise of AI Native Services

• The demand for AI native services is overwhelming, and companies like Anthropic, OpenAI, AWS, Microsoft, and Palantir are working on forward-deployed engineer models, with Anthropic creating a separate unit for this purpose, to meet the needs of enterprises 42s.
• The skill sets required for successful AI native services teams include an entrepreneurial spirit, flexible thinking, AI fluency, and a builder mindset, which is different from what traditional services businesses were built for, leading to the emergence of a new category of AI native services 2m6s.
• The engineering degree of difficulty for AI native services is very high, requiring experienced engineers who are generalist elite software engineers, with many being former founders, to juggle challenging technical problems and own solutions end-to-end 2m6s.
• Keeping teams small, with experienced engineers who can handle multiple challenges, is key to successful AI native services deployment, allowing a team of four to potentially have a $100 million impact on a business 2m6s.

Engineering Challenges and Team Structure in AI Native Services

• There is a trend of engineers being interested in deployment, rather than just working on the frontier of AI research, with many engineers wanting to work on transformative technology and make a real-world impact 2m6s.
• Engineers are drawn to working on AI native services because they see the potential for transformative technology to change the world, and they want to be part of that change, with two main career paths being to work on the technology itself or to work on deploying it 2m6s.
• Engineers are increasingly excited about two categories: building models and applications, or owning full business challenges end to end, and using technology to create new outcomes, which is a shift from the traditional excitement of joining a SaaS company 10s.
• The middle layer of SaaS companies is no longer as exciting to engineers, and there is a concern that these businesses may not be okay in the future, with the possibility of models subsuming some of their functions 42s.

Anthropic's Hiring Strategy and Talent Development

• Anthropic is not a traditional startup, having the backing of prominent investors like Blackstone, and is looking to acquire elite engineers, but is targeting applied AI talent rather than top researchers 2m6s.
• The company is facing a challenge in finding experienced applied AI talent, and is responding by hiring generalists and creating an environment for them to learn applied AI on the job, with hands-on experience and knowledge sharing rituals 2m6s.
• There is a concern that the next generation may have less generalist expertise, but Anthropic believes that AI is making it easier for people to become entrepreneurs and learn broad skills, and that their company is a good fit for people with these attributes 4m30s.
• The company is looking to hire people who can own problems end to end, have broad experiences, and are hungry to move the needle on something real, with over half of their team consisting of founders who have these skills 6m20s.
• Anthropic is not seeing a need to increase the supply of generalist expertise, but rather is focusing on creating an environment where generalists can learn and grow, and is optimistic about the potential for AI to enable entrepreneurship and skill-building 8m10s.

Challenges in Building a High-Performance Engineering Team

• The current environment presents challenges in building a super elite team of engineers, but it is not a new issue, as it has always been hard to build really talented teams, especially in the face of overwhelming demand 10s.
• The ambition is to become a scaled provider of AI native services that can meet the overwhelming demand of the market, with a goal of building a hundred billion dollar plus business, potentially even a trillion dollar company someday if executed well 2m6s.
• The key challenge is to go through the phase of hypergrowth without losing the emphasis on quality, and the company has a cultural value of over-delivering for clients, which is a key tension that is faced every day 2m6s.

Scaling and Maintaining Quality in AI Native Services

• To address this challenge, the company invests in things that will allow it to absorb new people faster, such as investing in technology, learning and development, to create a great experience for new team members and to over-deliver on projects 4m30s.
• AI changes the laws of physics for services businesses, bringing tons of efficiency internally, helping with oversight, consistency, and learning across projects, and enabling the creation of reusable building blocks to make engagements go faster 6m40s.
• The impact that can be had with a small team using AI is extremely high, unlike previous generation services businesses that required large teams, and this presents a significant opportunity for the company to make a substantial impact 8m20s.
• The opportunity to transform a whole business at scale is considered an AI phenomenon, where four-person engagements can have a significant impact, and while headcount needs to be scaled, it does not have to be excessively large, similar to a normal software company, 10s.

Future of AI in Services and Automation of Engineering Work

• The process of FDEs, or forward-deployed engineers, is already being automated to some extent, and it is imagined that an FDE agent could be created in the future, learning from the way employees help other companies implement AI workflows, 1m42s.
• Internal tools, such as Claude, are being used to assist with engagements and are expected to improve over time, creating room for larger outcomes and placing more value on human judgment, 2m6s.

Contact Information and Show Credits

• Listeners can connect with the company online by visiting their website at o.com, or by finding them on LinkedIn, and can also apply to be a full-time employee, or FTE, through the website, 4m30s.
• The show, Equity, is hosted by TechCrunch senior reporters, produced by Teresa Lo Concilio, and edited by Kell, with more information available on YouTube, LinkedIn, X, and Threads at Equity Pod, and on the TechCrunch website at techcrunch.com/events, 5m40s.