The FDE Playbook for AI Startups with Bob McGrew

What is a Forward Deployed Engineer? [1:57]

• An FDE is typically a technical engineer who works at the customer's site.
• Their role is to bridge the gap between what the product can do and what the customer actually needs.
• They tackle problems that may not have been solved before, believing they can solve them for a specific customer with effort.
• FDEs deliver valuable outcomes by adapting the existing product, figuring out how to build use cases, and ensuring the software works for the customer.

A forward deployed engineer is someone typically technical and engineer who sits at the customer site and fills the gap between what the product does and what the customer needs.

The Forward Deployed Engineer (FDE) is a technical professional embedded at a client's location to adapt a product to their specific requirements. This role is vital for addressing novel problems and ensuring the software effectively meets the customer's unique needs, thereby delivering significant value.

Palantir's Origins and the Birth of the FDE Strategy [4:40]

• Palantir initially focused on building software for the intelligence community, which presented challenges due to the secretive nature of potential users.
• They adopted an unusual approach: building a demo and taking it to potential customers, who would then provide feedback on how it could be improved.
• This involved actively listening to customer needs and making modifications, akin to the modern "get out of the building" founder advice.
• The FDE strategy emerged because their customer base (intelligence agencies, law enforcement, military) had unique needs, making a one-size-fits-all product impractical.
• Instead of building separate products for each client, they developed a customizable platform.

The interesting way to think about the beginning of Palunteer is that uh when we got started, the focus of our company was to build software for the intelligence community, specifically software for spies.

Palantir's early days were defined by the challenge of developing software for a clandestine community. Their innovative solution involved a customer-centric, demo-driven process to understand and meet diverse user requirements, leading to the creation of a flexible platform rather than rigid, single-use products.

The FDE Model: Product Discovery and Customer Expansion [11:30]

• The FDE's role evolved to be instrumental in product discovery. They would implement the existing product, identify gaps, and essentially "pave the road" for product development.
• The product and engineering teams would then analyze these FDE-driven customizations to generalize them for broader application across multiple customers.
• This contrasts with sales-led product discovery, where external sales teams relay information. FDEs offer a deeper, internal perspective.
• The FDE approach enables a "land and expand" strategy, moving beyond initial problem-solving to identify and address more valuable enterprise-wide issues.

What he realized we needed was for the FTEEs to act as product discovery. So they would go to the site, they would take the product as it was and they would fill the gap between what the product did and what the users needed.

FDEs are central to uncovering how a product can best serve customers by actively working within their environments. This direct engagement provides invaluable insights for product teams to generalize and improve the software, leading to deeper customer penetration and value.

Palantir's FDE Team Structure and Hiring Philosophy [25:46]

• Palantir's FDE model relies on two key roles: the "echo team" and the "delta team."
• The echo team consists of embedded analysts who speak with users, identify key problems and use cases, and manage customer relationships.
• The delta team comprises highly skilled software engineers who rapidly build and deploy solutions based on the echo team's findings.
• Echo team members often come from the customer's domain (e.g., former military officers) but must also be "rebels" or "heretics" who recognize existing systems' limitations.
• Delta team members are strong prototypers, capable of delivering outcomes quickly, even if the initial code isn't perfectly polished for long-term maintenance.

And they need to be someone who understands how things are done right now and recognizes that it's insufficient, that it doesn\'t work.

The success of Palantir's FDE strategy hinges on distinct team roles: domain experts identifying customer needs and engineers rapidly building solutions. Crucially, team members must possess a critical perspective on existing systems and a drive for innovation to facilitate significant technological advancements.

FDE Model vs. Consulting: Distinguishing Value [38:14]

• A common misconception is that the FDE model is simply consulting in disguise.
• The key difference lies in the business model and scalability. Early FDE engagements might appear to lose money but are intended to refine the product for future scalability.
• As the product improves and becomes better suited to customer needs, the need for extensive on-site teams decreases, and profit margins increase.
• The FDE model aims to earn the "right" to tackle more significant problems, demonstrating repeatable value and a decreasing cost per outcome over time.

From a business model perspective, one of the key things that you will see that you should see is that it may be the case that your when you go into you do a new deployment at a customer that you\'re actually losing money early on.

While superficially similar to consulting, the FDE model is fundamentally a product development strategy. It involves an initial investment in understanding and adapting to customer needs, which ultimately leads to a more refined, scalable, and profitable product, differentiating it from pure service-based engagements.

The Role of the Product Team in the FDE Model [44:44]

• The product team's role shifts from building highly verticalized products to holding a broader product vision.
• They must identify the generalizable version of problems discovered by FDEs that can apply to multiple customers.
• A failure mode is directly incorporating a single customer's specific solution into the product, making it over-specialized.
• The product team needs to think at a higher level of abstraction, understanding how to generalize solutions within the context of the overall product ontology.
• This requires product managers who can conceptualize solutions that work across various customer needs and market segments.

The role of the product team is is really to hold the product vision and so you have to think when I see this new problem that we\'re seeing at a customer site what is the generalizable version of this that applies to the next 10 customers

In the FDE model, the product team acts as a curator of innovation, translating individual customer solutions into broadly applicable product features. Their focus is on generalization and abstraction, ensuring that the core product evolves to meet diverse needs without becoming overly specialized for any single client.

The Palantir Ontology and Generalization [52:32]

• The development of Palantir's ontology is a prime example of the FDE model driving product generalization.
• Initially, they considered specific database tables for distinct entities like people and money.
• However, they realized this approach was not scalable for multiple customers and opted for a higher level of generalization.
• The ontology allows for customer-defined objects, properties, and relationships, enabling a flexible data structure.
• This higher-level abstraction allows for common operations to be applied across different types of entities, rather than narrowly defining them per customer.

And instead of thinking about specific types of objects um we should allow that to be defined per customer by the forward deployed engineering team.

The Palantir ontology exemplifies how the FDE model fosters product generalization. By abstracting concepts to a higher level, the system can accommodate diverse customer-specific definitions and needs, creating a powerful and adaptable platform rather than a collection of siloed solutions.

FDEs and Product Teams: Navigating Tension and Incentives [1:02:34]

• Tension exists between FDEs focused on immediate customer needs and product teams aiming for generalization and maintainability.
• FDEs are incentivized to solve specific, pressing problems efficiently, which might lead to less generalized solutions.
• The product team, conversely, must consider how to make solutions work across multiple customers and ensure long-term product health.
• Early product development often involved FDEs building solutions, with the product team then generalizing them, sometimes with input from FDEs at other customer sites.
• This collaborative process helps align incentives and ensures that generalized features meet the needs of both new and existing customers.

The incentives are different. And so, you know, you\'re solving one very particular problem and it makes a lot of sense to just take the simplest approach to solve that problem.

The dynamic between FDEs and product teams involves managing differing incentives and perspectives. FDEs prioritize immediate customer solutions, while product teams focus on broader applicability. This inherent tension, when managed effectively through collaboration and shared understanding, drives the development of robust and valuable products.

The AI Agent Market and the Rise of the FDE Model [1:20:16]

• The FDE model, once niche to Palantir, is now exploding in the AI agent market.
• This adoption is surprising, as it was previously seen as a unique strategy for companies like Palantir selling to governments.
• The FDE model is taking off because AI agents represent a new market category with no dominant incumbent product.
• Significant product discovery is required to understand what AI agents can and should do.
• This discovery process is best done from within the enterprise, making the FDE model highly effective.

With AI agents, there is no incumbent product. And so that I think is why you\'re seeing the FTE model taking off because there\'s so much product discovery to do.

The current surge in AI agent companies adopting the FDE model is driven by the lack of established players in this nascent market. The necessity for extensive product discovery, understanding unique customer needs, and building novel applications makes the FDE approach an ideal strategy for this evolving landscape.

Misconceptions and Pitfalls of the FDE Model [1:30:51]

• A common mistake is for startups to treat FDE work as pure consulting, failing to generalize learnings back into the product.
• Pricing is a critical differentiator: FDE is about selling an outcome, not just software installation. Pricing should reflect the value delivered.
• Traditional SAS pricing (usage, subscription, seats) doesn't apply; FDE pricing should be outcome-based and often leads to larger, more flexible contracts.
• Startups need to avoid building products solely based on customer requests that might not be truly impactful.
• Securing executive buy-in is crucial, as FDE projects often need to align with top leadership priorities to overcome organizational inertia and IT hurdles.

Fundamentally what you\'re selling with the FD model is that you\'re not selling the installation of software. You\'re selling an outcome.

A key pitfall in adopting the FDE model is mistaking it for consulting and failing to monetize outcomes effectively. Startups must understand that they are selling solutions, not just software, and price accordingly. Moreover, securing executive sponsorship is vital for navigating complex enterprise environments and ensuring project success.

The Palantir FDE Model: A Learning Company [1:46:19]

• The FDE model fosters a culture of continuous learning within a company.
• Unlike established giants like Google or Meta, which can coast on existing success, younger companies must constantly adapt and learn.
• Palantir, despite its size, remains a "learning company" because it continues to operate with the agility and problem-solving ethos of a startup.
• This environment cultivates skills essential for founders, explaining why Palantir has spun off numerous successful startups.
• The FDE process involves constant iteration, learning from customer interactions, and refining both product and approach.

The kind of company you have to build is a learning company and I think everybody wants to build a learning company but if you\'re a company like Google or Meta it\'s very easy not to learn because what you\'re doing right now was working

The FDE model cultivates a "learning company" culture, where continuous adaptation and innovation are paramount. This environment, exemplified by Palantir even as a large organization, fosters the skills and mindset necessary for entrepreneurial success, leading to the incubation of new ventures.