So does everyone else.
In the last few months, some of the biggest names in AI have announced new partner programs to build out and support AI-forward services partners.
- Google announced a $750 million investment to accelerate partners’ agentic efforts
- Anthropic announced a $100 million investment in its Claude Partner Network…for 2026
- OpenAI announced its Frontier Alliance partner program alongside a who’s who of GSIs
They aren’t the only ones. Microsoft continues to throw investment behind its AI Cloud Partner Program. Palantir and the GSIs have been announcing partnerships left, right and center. Other AI players like Writer.ai are redesigning their programs, while larger SaaS players like ServiceNow and Salesforce are building specific agentic and Forward Deployed Engineer (FDE) programs.
So much for services being dead!
Yes, the business model for tech services is undergoing major changes. However, these moves highlight that traditional and AI vendors still see services partners as a critical component of their growth strategy. Especially if they want to make traction in the enterprise.
They need trusted partners that can upskill and train a growing number of clients on how to use these new tools in the right way. Partners that understand how to navigate enterprise complexity, and can develop and deploy use cases that drive consumption and results. Partners that can build, govern and secure agentic systems at scale.
For the last few years, the newer AI platform vendors like Anthropic and OpenAI have focused on relationships with the largest GSIs and a small handful of AI specialists. However, as models evolve and AI gains traction beyond early adopters, they’re now looking to expand their partner networks to a wider group and throwing significant investment behind it.
And that’s good news for services firms looking to play in the AI era.
What AI Platform Vendors Are Looking For in Their Partners
When we speak with channel leaders at these vendors, what becomes clear is that they’re looking for quality over quantity in their partner programs. These teams are being inundated with requests from potential partners – a mix of small consultancies looking to build a practice, AI natives building from the ground up, and larger traditional systems integrators looking for their next big growth partner.
One channel leader said they’re following the 90/10 rule…doing what they can to get through the noise to separate the 10% of partners who really understand how to operate in this new environment. Those who are adept with technologies that are advancing every day, and that aren’t hooked on implementations and traditional T&M projects.
Here are the most common areas we hear from the AI players on where they need help:
- Internal AI transformation (process/organizational change)
- Use case development and execution
- LLM deployment and customization
- Data transformation/normalization for AI
- Training and activation
- AI governance and security
It’s that last area that continues to come up in more recent conversations.
A Different Channel Playbook
While these programs may seem similar to the SaaS and software partner programs of the past, when we speak with senior services leaders who have had early success with the AI platform vendors, a few key themes emerge.
Less Lead Flow, More Co-Delivery
Don’t expect leads to rain down from vendors. It’s more of a co-delivery motion.
Hyperscalers and traditional software firms have spent years building sales and support teams. Even after recent layoffs, they still have large teams of salespeople and client managers dedicated to drumming up deals and pulling in partners. That is not the case at companies like Anthropic and OpenAI. Their sales and partner teams are much smaller than those of companies that have been around for decades.
For service providers, this means the lead flow playbook is being replaced by a model in which they must bring their own opportunities to the table, while proving they have the technical chops to deliver on their promises.
For service providers, this means the lead flow playbook is being replaced by a model in which they must bring their own opportunities to the table, while proving they have the technical chops to deliver on their promises.
Randall Hunt, CTO at Caylent, a Premier tier AWS Services partner that established a partnership with Anthropic back in 2024, expands on this:
“For us, Anthropic is a strategic relationship, but it’s not a massively strong channel sales model compared to the hyperscalers. It’s more of a co-delivery model. We interact with the Anthropic team frequently because so many of our customers use their services.”
Those interactions could be around large complex engagements or as simple as asking for an increase in tokens or help troubleshooting. All are valuable.
For example, Caylent recently built an automation for a municipal infrastructure company in the water and utilities sector that analyzes maps and documents, scopes the project and calculates the cost of materials for a bid. But there was a problem: Caylent engineers kept getting a weird safety error from Anthropic with some of the maps. As it turned out, Anthropic’s safety algorithm filtered the word “manhole” on maps. “When we finally figured out what was triggering it, it was quite a funny moment,” Hunt said.
Here, the co-delivery relationship enabled Caylent to point out the legitimacy of its map-analysis use case and ask the model provider to address the issue. Anthropic fixed the problem in less than 24 hours. Hunt noted that even without the relationship, Anthropic would have responded, but without the partnership, they would have had to go through Anthropic’s standard support channels, which might have taken days.
More Domain Experts, Less Generalists
Vendors have always looked for domain experience in services partners, but now it’s a requirement.
AI vendors have to be the experts in their technology, which is advancing every day. They need partners that identify and apply that technology where it can create real business value inside complex, regulated, legacy environments. That requires data and process fluency and domain judgment — not just model/API expertise.
One of the channel partners we spoke with said they are seeing some of the most exciting use cases coming from vertical specialists who aren’t necessarily deep, technical AI experts. Vertical and domain specialists know customers’ problems deeply and are better equipped to translate AI into high-value workflows, ultimately putting enterprises on the path to greater consumption.
Vertical and domain specialists know customers’ problems deeply and are better equipped to translate AI into high-value workflows, ultimately putting enterprises on the path to greater consumption.
Vertical or domain specialists are also well-positioned to drive consumption through repeatable, packaged industry use cases. Proven reference architectures and productized IP can be built when you see the same problem over and over again, and it can help customers move forward with more certainty than bespoke generalist work.
Finally, enterprise AI adoption increasingly hinges on governance, security, compliance, data quality, and change management. Domain-focused firms are better positioned to understand industry-specific risks and build buyer confidence, especially in sectors such as financial services, healthcare, utilities, and other regulated markets.
Less Analysts, More Architects
AI leaders aren’t looking for resellers, referral partners, or armies of implementers. They need partners that understand and can navigate complex environments that come with legacy systems, messy data, AI agent sprawl, and teams that may not be ready for what’s to come.
Succeeding in these environments requires an architect and product mindset.
Succeeding in these environments requires an architect and product mindset. It’s why we’re seeing services firms hiring fewer general Business Analyst and Project Manager profiles, and more AI Solutions Architects and Forward Deployed Engineers. People who can act as a single translation layer between executives, engineers and users, and leverage AI for task completion.
These are the people who can determine what is worth building, prototype quickly with real data and APIs, and navigate constraints like compliance, fragmented data, legacy infrastructure, cost ceilings and organizational politics.
It’s why GSIs have had such a leg up here. However, the AI platform vendors can’t only rely on the GSIs if they want to scale and continue to capture share. It’s why they are opening up to other types of partners who can pair business credibility with technical chops, and move faster than the GSIs.
Tribe AI is a great example of this new kind of partner. They have become a go-to partner for the likes of Anthropic and Google because their experts are technical enough to build, strategic enough to advise, and practical enough to get something adopted inside the messy reality of the enterprise.
The AI platform vendors are deploying their own FDEs alongside partners to surface product gaps and ensure the technology gets used successfully, but today these teams are relatively small. For example, Anthropic — a company with an estimated $30 billion ARR — has a 200-person Applied AI team. They expect to scale to 500 by the end of 2026, but they know they need to keep the bar high.
While it might seem like these captive FDE teams would compete with partners for customers, the demand for these skills is so high and the skillset is so nascent, there is plenty of work to go around.
Let the Games Begin
As you can imagine, service partners are flocking like bees to honey to these programs – as they should.
The early standouts in these programs might have been the GSIs, who have invested billions into arming consultants with the latest models, retraining or acquiring AI-native talent and building out IP. But a rapidly changing and unsettled market creates a more even playing field for new players, and the investments in these partner programs show that AI vendors are ready to expand.
A rapidly changing and unsettled market creates a more even playing field for new players, and the investments in these partner programs show that AI vendors are ready to expand.
The first step is to register for these programs and start getting certified. But what does it take to really stand out and be a player?
Prove your value. Don’t expect instant leadflow from the AI vendors. Remember, these vendors are still trying to sort through the 90/10 rule and figure out which partners they can trust. So bring your own customers to the table. Be armed with case studies and a clear case for where you add value. Show how you are using their models and products in your own business.
Come with a point of view. Demonstrate that your domain expertise is real. Be vocal and confident in your opinions on where AI and agents can make a meaningful impact, and how to get predictable performance. Market yourself with data-backed thought leadership and proven case studies that only you have – not those that AI spits out.
Lead with IP. Don’t come to the table with buzzwords and slideware. Come with pre-built agents or productized IP that shows you know what you’re doing, and can accelerate results and reduce risk for both customers and the vendor.
Mirror the model. Understand how Applied AI or FDE teams inside these AI platform companies operate. Start shifting from T&M to more outcome-based contracting. Vendors might not be overly prescriptive out of the gate about having to work in that fashion, but if you do, you’ll certainly get more notice.
Be adaptable. Recognize that the partner programs for some of the AI players are still developing. They aren’t as mature as the hyperscaler or software partner programs that are nearly 15 years old. There will be growing pains as these programs evolve, so be flexible and open to new ways of working.
In the end, the winners will be those that can show up with real customer opportunities, credible industry and process knowledge, AI-native technical talent, reusable patterns, and the ability to help customers move from interesting demos to consumption-driving systems.