Cognition acquires Poke: AI agent personality matters

Timeline of the deal: what happened and why now

On July 24, 2026, TechCrunch reported the acquisition of Poke by Cognition . Cognition is the company that developed Devin, the coding agent that in 2024 had caused a stir for its autonomous code writing and debugging capabilities. Poke, on the other hand, had specialized in a different segment: the conversational style of AIs and the user interaction model.

However, the deal wasn't announced as a simple talent acquisition. Cognition explicitly stated its intention to integrate Poke's conversational model directly into Devin. Therefore, it's a strategic product choice, not just a defensive move in the AI talent market.

So, the real question for digital professionals isn't who bought who . The question is: why did a coding agent company feel the need to acquire conversational UX skills?

The underlying model: when personality becomes architecture

For years, the debate around AI agents has focused on benchmarks. Which model answers most accurately? Which has the widest context window? Which costs less per token? These metrics remain relevant. However, the market is signaling that they are no longer enough to differentiate a product.

In fact, Gartner identified conversational UX as one of the critical factors in enterprise adoption of AI agents . Users abandon technically superior tools if the interaction feels cold, ambiguous, or inconsistent. On the other hand, agents with a recognizable personality and a tone suited to the context build greater trust and ongoing adoption.

This is not a new phenomenon in software. Much like what happened with graphical interfaces in the nineties—where UX transformed the personal computer from a tool for techs into a mass product—conversational UX is now redefining the accessibility of AI agents. Cognition figured this out before many competitors.

Devin before and after Poke: what actually changes

Devin was already a capable tool. It could write code, run tests, open pull requests, and browse technical documentation on its own. In particular, it was appreciated by dev teams for its ability to handle complex tasks without continuous supervision.

Despite this, some recurring feedback concerned the quality of the interaction: overly verbose responses, uniform tone regardless of context, lack of conversational cues to help the user understand the task status. Poke had worked on exactly these aspects. Therefore, the integration does not add functionality in the strict sense: it improves the perceived quality of the experience.

For digital managers evaluating the adoption of coding agents in their teams, this has practical implications. An agent with a polished conversational UX reduces onboarding time, decreases misunderstanding errors, and increases the team's trust in the tool. Consequently, the ROI of adoption improves — not because of the model's capabilities, but because of the user experience.

Who wins and who loses in this scenario

The acquisition redraws competitive priorities in the AI agent market. Let's look at the main effects.

• Cognition strengthens the moat: integrating the conversational personality into Devin makes it harder for competitors to replicate the user experience, even with the same underlying model. UX is harder to copy than a technical architecture.
• Generalist competitors are under pressure: tools like GitHub Copilot or Cursor mainly compete on features and IDE integration. Also, they now need to consider the conversational dimension as a differentiator.
• AI personality startups are gaining visibility: the buyout shows this category has real market value. Because of that, we expect to see more deals like this in the next 12 to 18 months.
• End users benefit from the competition: competitive pressure on conversational UX will produce more usable tools for everyone, including the tech teams of Italian SMEs.

SHM Studio's take: AI personalities and corporate brand identity

From SHM Studio , we observe this dynamic from a specific perspective: that of Italian companies integrating AI into their marketing, communication, and product development processes. In fact, the issue of AI agent personality does not only concern developer tools. It concerns any conversational interface that a company exposes to its customers or its teams.

Think about customer service chatbots, assistants integrated into CRMs, tools for AI applied to marketing . In all these contexts, the agent's tone communicates — whether it wants to or not — something about the brand. An AI assistant that replies in a generic, bureaucratic, or inconsistent way projects a sense of distance. On the flip side, an agent with a recognizable voice that matches company values strengthens the brand identity even during everyday tasks.

So, the question marketing leaders need to ask isn't just «does this AI tool work?». The real question is: «does this AI tool sounds like our company?». This distinction is going to matter more and more as AI agents show up in customer touchpoints.

Operational implications for Italian companies

Translating this acquisition into concrete actions requires a shift in perspective when evaluating AI tools. Here are some practical tips for digital and marketing managers.

• Include conversational UX in your selection criteria: when evaluating an AI agent — for internal use or customer-facing applications — adding qualitative metrics: tone consistency, clarity in ambiguous answers, and error handling. These aspects impact real-world adoption much more than many technical benchmarks.
• Defining a company "AI voice": much like brand guidelines for copywriting, it is useful to define guidelines for the tone of corporate AI agents. This is especially relevant for anyone managing chatbots or assistants across customer touchpoints. Our team of strategic copywriting can help with this definition.
• Evaluate the customizability of tools: not all AI agents allow you to configure tone and style. Therefore, during procurement phases, check the level of customization available — especially for companies with a strong brand identity.
• Monitor market trends: the acquisition of Poke is likely the first in a series. In the coming months, we expect other players to integrate conversational UX skills. Therefore, those planning investments in AI agents should follow this evolution before consolidating their technological choices.

For those who manage digital marketing strategies or LinkedIn campaigns , AI personality also has a content dimension: the tools used to generate copy, respond to leads, or manage comments must be consistent with the brand voice. Similarly, those who use AI to optimize google ads campaigns must verify that the generated texts respect the company's communication tone.

The still open worksite: what we don't know

The acquisition is recent and many details remain to be clarified. In particular, it is not yet clear how Cognition will technically integrate Poke's conversational model into Devin. Furthermore, we don't know if this personality will be customizable by enterprise users or if it will be a fixed product setting.

Harvard Business Review analyzed how trust in AI systems depends largely on the consistency of the interaction . However, building consistency on a large scale — with different users, different tasks, and different contexts — is a non-trivial engineering challenge. Poke had developed interesting approaches in this direction, but their scalability within a product like Devin remains to be demonstrated.

Finally, the question of long-term differentiation remains open. If AI personality becomes a standard competitive axis, Cognition's advantage could erode quickly. Therefore, acquiring Poke is a step forward, but not necessarily a sustainable advantage over time without continuous investment in this direction.

Next moves: what to watch in the coming months

For digital and marketing leaders wanting to stay ahead of AI trends, we suggest keeping an eye on a few key signals.

• Other acquisitions in the conversational UX segment: if other major players follow Cognition, the signal will become a consolidated trend. Besides this, a new category of vendors specialized in "AI personality layer" could emerge.
• Product updates from competitors: GitHub Copilot, Cursor, and similar tools might respond with tone personalization features. So, it's worth keeping an eye on release notes over the next few quarters.
• Enterprise user feedback on Devin post-integration: the first user experience reports after integrating Poke will give us a real-world sign of how conversational UX actually impacts adoption.

To learn more about how to integrate AI tools into your company's marketing and communication processes, the team at SHM Studio is available for an initial consultation. You can contact us directly or explore our areas of expertise on web development , SEO and insights from the blog . Plus, if you're trying to figure out where your company fits in the AI world, you can find some handy tips in our section dedicated to AI services .