John Fries

The hottest role in AI startups right now isn’t Forward Deployed Engineers. It isn't GTM Engineers. It’s Deployment Strategists. Decagon calls it an “Agent Product Manager.” Harvey calls it a “Solutions Architect.” Palantir Technologies has had versions of this role for years. And the salaries are climbing fast: - Decagon: $200k–$285k - Palantir Technologies: $120k–$200k - Figma: $150k–$260k - Ramp: $100k–$180k - Harvey: $190k–$260k So who are these people? They are usually pseudo-technical -- CS or engineering majors, or folks with technical work experience. Many come from 2 years in consulting, IB, or PE, then jump into startups to get their hands dirty. They are young, hungry, polished, and comfortable being in front of customers. What do they actually do? They make sure enterprise AI deployments succeed. A $100k+ deal does not survive on a nice pitch or a self-serve onboarding flow. It survives if the customer sees value in the pilot. That means: - Embedding directly with the customer - Designing prompt logic for specific workflows - Working with engineering to align integrations and data flow - Helping exec teams define their AI roadmap - Running feedback loops into product and GTM Why does this role matter so much? Because enterprise AI is messy. Integrations, data transfer, and adoption make or break a deal. Most buyers are using AI for the first time, and each has unique workflows. Deployment Strategists bridge that gap. They own the outcome. They are accountable for making pilots successful, which often means millions in revenue down the line. At Conversion, Sam Bochner has been leading this work for us. We are now thinking about scaling it into a full team. Because a few successful pilots can fund an entire department, and the cost of failed deployments is too high to ignore. Is this just a rebrand of customer success? Not really. Success is about answering tickets and renewals. Deployment Strategy is about going deep with a few enterprise accounts, extracting maximum value, and ensuring the pilot closes into a multi-year contract. Call it Agent PM, Solutions Architect, or Deployment Strategist. Whatever the title, this is becoming one of the most important roles in AI SaaS.

602

90 Comments

A bit late to the party, but Brett Queener's post on the death of the schema from earlier this month is one of the best pieces of writing out there on the future of SaaS x AI. Paralleling, in a way, Andrew Karpathy's most recent YC Startup School talk on the future of programming, Brett talks about the nerdiest, yet most vital leg of the stool that traditional enterprise software stands on - the database schema. A schema is fundamentally a way to model a business process (e.g., CRM, HCM, ERP) into a data model (e.g., a database). Schemas are why we use terms like "Accounts" and "Opportunities" so fluidly. But they're also why SaaS can sometimes feel rigid. But with AI, schemas don't have to be fixed anymore. And because schemas aren't fixed, UIs don't need to be either. Brett hints at the big prize at the end of the post - Reinforcement Learning for enterprise domains. While RL has been optimized (through Human Feedback) for our general use of chatbots like ChatGPT and has been further tuned for technical fields like math and coding, it still has a ways to go in being optimized for vertical and domain-specific use cases. Dynamic schemas + contextualized UIs + domain-specific RL = software that will be more powerful than any of us can imagine. It's not coming overnight, but it will be here soon. Link to the post is in the comments. Definitely worth a read. Thanks to Brett for writing it!

137

24 Comments

A fintech giant wanted us to build custom features for their deal. My engineering brain said "absolutely not." That custom work became the path to our first $1M+ contracts. A lesson every B2B founder needs to hear: Early at Assembled, we were two engineers and one salesperson, Brian Sze. Brian came in with a $100k deal, With one of the fastest growing fintechs of the 2020s. Now public. "They want us to do all this custom stuff," he said. My immediate reaction: "That doesn't fit our model. We shouldn't do work that's not really what we want to do." Classic engineering mindset. Build the perfect product. Make customers adapt to it. Don't pollute your clean architecture with one-off requests. Brian looked at us like we were insane. We had a heated debate. He wrote up a whole decision document on why this was an exception. Thank god we listened. This giant became our customer. In 2020, Bitcoin and meme stocks exploded. This customer went from hundreds to thousands of support agents overnight. We woke up every morning checking dashboards. Revenue going up and to the right. Usage going up and to the right. Everything growing because we were part of this fintech’s growth story. But here's the deeper lesson: That "custom" work wasn't actually custom. It was vanguard work for the next tier of enterprise customers. Every time we had an order of magnitude contract increase ($100K, $1M, $10M) it started with custom work. But that custom work became the template for the next several deals at that level. The million-dollar custom features became the standard for all million-dollar customers. Your best customers pull you into the future. But only if you're willing to schlep. The engineering mindset says "build the perfect product, then scale it." The winning mindset says "build what your best customer needs, then generalize it." Don't confuse purity with strategy. Sometimes the messiest path leads to the biggest outcomes. We almost said no to custom work that defined our entire business model. The difference between a consulting shop and a $100M company isn't avoiding custom work. It's choosing the right custom work from the right customers.

87

7 Comments

Launch day! Continuing the drum beat on new provider integrations: today we're announcing a native integration with Anyscale! Anyscale kicks off a series of AI-specific provider integrations we're going to be announcing for the remainder of the year. When I started Vantage, I had a belief that companies were going to be using more and more third-party providers beyond the main three hyperscalers to run their infrastructure. Today, AI has materially accelerated that trend. Each of these AI providers has their own bespoke pricing structure that we need to work with to provide the single pane of glass view into total cost of ownership for an organization. As companies leverage these AI providers, they expect that their Finops platform has support accordingly. As with the ~25 other native integrations Vantage currently supports, Anyscale costs will support all the #FinOps features you know and love across cost allocation, budgeting, forecasting and beyond. Getting starting is as simple as providing an API key and we do all the heavy lifting from there. To see all the details, check out the blog post below.

55

1 Comment