.webp)
The latest episode of Presales Podcast featured a fascinating conversation with Akash Ganapathi, CEO of Opine, who brought a unique perspective on how AI strategies should differ between startups and enterprises. Hosted by Jack Cochran and Matthew James, the episode explored why one-size-fits-all AI approaches don't work and what the future holds for presales professionals.
Listen to the full episode now!
The Great AI Strategy Divide
One of the most compelling insights from Akash was how fundamentally different AI implementation should be between startups and enterprises. "The big difference to me is the specialization of roles," he explained. "In large organizations, there's deep specialization... where at a startup, it's just like an everyman, really quite general, wearing multiple hats."
This difference in role specialization creates a paradox in AI strategy. While a large public company might give enterprise AI licenses to thousands of employees (potentially creating redundant efforts), that same bottom-up approach could be exactly what a startup needs to foster innovation and discover unexpected use cases.
Debunking the AI Replacement Myth
Akash was particularly passionate about addressing what he sees as harmful marketing around AI replacing sales engineers entirely. "I'd say the biggest myth that I'm hearing just consistently across presales, whether it's a startup or an enterprise, is just this idea of the AI SE that's basically going to replace the SE function."
The reality, according to Akash, is that these tools fundamentally misunderstand what SEs actually do. "These tools are basically just knowledge base question and answer chats," he noted. "I think that's almost no SE's job description is just answering technical questions. There's so much more to the role with respect to putting together a solution, having a value narrative that you're conveying, building a relationship with technical stakeholders."
The Current State of AI Limitations
The conversation took a technical turn as Ganapathi explained the current limitations of large language models. Context windows, even in the best models today, are limited to about a million tokens (roughly equivalent to a few large books). While this might seem substantial, it pales in comparison to the vast knowledge bases that enterprise organizations accumulate.
"When you start thinking about all of the context of my customers and my prospects and all their use cases and everything I've ever spoken about across my entire revenue org, suddenly that context is super massive," he explained. This limitation means that Retrieval Augmented Generation (RAG) systems, which attempt to provide additional context, achieve only about 75% accuracy compared to the 99.9% accuracy within the context window.
The Human Element Remains Critical
Jack Cochran drew a compelling parallel to personal computers in the 1980s, noting that despite predictions of job displacement, computing technology ultimately created more jobs than it eliminated. "As long as the people are buying, they're going to want to buy from people," he observed.
Akash agreed, emphasizing that "as long as the buyers are people, the sellers are people, that's what it boils down to for me personally." The focus should be on using AI to handle the administrative tasks that SEs don't enjoy such as Salesforce updates and first drafts of statements of work. This frees up time for the relationship-building and strategic thinking that drew them to the role in the first place.
Looking Toward 2026-2027
When asked about the future, Akash painted a picture of presales teams managing significantly more revenue per team member. "I think we're headed down a lot of the themes that we've discussed so far. Revenue teams are managing a lot more revenue per team member. That would be true for presales orgs as well as AEs."
He also predicted a shift toward hiring less experienced SEs who can be quickly enabled through AI tools, particularly in larger enterprises. However, he was optimistic about the core of the role: "On the more optimistic side, I think presales will become this very strategic role and they'll have a lot less minutiae to deal with in our roles, a lot less project management, a lot less updates and status and documents."
Practical Recommendations
For mid-market companies looking to implement AI strategies, Akash recommended taking a more enterprise-like, forward-looking approach rather than the experimental startup model. The key is to think beyond just the current team: "How do we enable not just the current team, but the next X hires on the team, how do we take all of this tremendous knowledge and value that the SE org and presales org produces in general and enable other parts of the revenue org?"
He emphasized the importance of top-down strategic thinking while still gathering input from team members: "You should always consider the inputs of everyone on your team... But if you're relying on the execution to be bottoms up, someone's going to build the process automation for meeting prep, someone's going to build the thing for account research... that is going to turn into super redundant effort."
The Opportunity Ahead
Akash concluded with an encouraging message for presales professionals: "I think the most exciting thing you could do as an SE or an SE leader right now is be the one owning the AI strategy for presales in your organization... There's an opportunity right now where every revenue org is trying to figure out what their AI strategy is, and presales is a critical part of that strategy."
Rather than viewing AI as a threat, Akash sees it as an opportunity for presales professionals to become even more strategic and valuable within their organizations. By embracing AI as a tool for handling administrative tasks, SEs can focus on what they do best: building relationships, solving complex problems, and driving revenue growth.
You can listen to all episodes of the Presales Podcast by visiting our website at presalescollective.com/podcast. Join the Presales Collective Slack community at presalescollective.com/slack to stay updated on future episodes and join in the conversation.
