Over the past year, Aquiva team members showed up across the Salesforce ecosystem in many forms. Speaking, listening, and learning alongside customers, partners, and product teams.
What stood out was not any single event or presentation, but the consistency of the conversations that surfaced across them. Questions about how agentic systems behave in real environments. How teams move from experimentation to production. Where trust, governance, and human judgment still matter. And, how industry and human connection shape whether any of this delivers value.
This is not a comprehensive recap of every stop along the way. It is a reflection on the topics that kept resonating across the ecosystem.
1. Agentic AI is moving from promise to production reality
Across the Salesforce ecosystem, conversations about agentic AI are maturing quickly. Among experienced practitioners, the focus has shifted from experimentation to execution. The question is no longer what agents can do, but how to design them so they hold up in real environments.
That theme was reinforced through AI Agents from the Trenches, presented by Robert Sösemann in Dublin, Porto, and Dubai.
The sessions centered on a clear point of view. Agentic systems can reduce UI complexity and replace brittle glue code with reasoning-driven orchestration, but only when they are built with intention. Conversational interfaces, guardrails, testing, and clear instructions are foundational. Without them, agents simply move complexity elsewhere. The signal across venues was steady and unambiguous: delivering agentic AI at scale requires more discipline, not less.
2. Building with Agentforce requires discipline, not just capability
As Agentforce adoption accelerates, a consistent set of questions keeps surfacing among experienced Salesforce practitioners. Not about whether Agentforce works, but about how to design agents that are reliable, scalable, and usable over time.
Those questions were addressed through Top 10 Best Practices for Building with Agentforce, delivered by Aquiva leaders across multiple venues: Greg Wasowski at TDX San Francisco, Michael Holt at World Tour London, and Alex Bariyev at World Tour New York.
The guidance centered on restraint and intent. Start with narrowly defined use cases. Iterate before expanding scope. Be explicit in instructions and topics. Design for variability in user input and data availability. Embed security, governance, and human oversight from the beginning. Adoption and change management were treated as architectural concerns, not afterthoughts. The recurring insight was clear: Agentforce success is less about model sophistication and more about disciplined system design that respects both technical and human constraints.
3. Trust, security, and governance are foundational design concerns
As agentic systems move closer to production, conversations across the ecosystem consistently returned to trust. Not as an abstract principle, but as a design requirement that shows up in architecture, tooling, and review processes.
That perspective was central to Secure App Development for Agentforce, co-presented by Greg Wasowski alongside Salesforce at TDX San Francisco, and reinforced in sessions focused in Greg’s Dreamforce presentation, Agent Decisions: Using Variables and Filters in Agentforce.
Security review, static code analysis, and guardrails were not framed as friction. They were positioned as enablers of adoption. Variables provide structured context that anchors agent behavior in known data and business rules. Filters act as deterministic controls that make outputs predictable and auditable. Together, they reduce risk while increasing confidence in agent-driven workflows.
What resonated most was the shift in mindset. Trust is not earned after deployment. It is designed into systems from the beginning, through disciplined engineering practices that make agentic behavior observable, testable, and accountable.
4. The agentic enterprise conversation is shifting toward responsibility
Beyond implementation details, a broader conversation is taking shape around what it means to operate agentic systems inside complex organizations.
That discussion came into focus during the Limitless Lab panel, “Aquiva Limitless Lab Social: AI, Unpacked — Real-World Guidance for Every Stage,” moderated by Greg Wasowski, with perspectives from leaders across Salesforce, ISVs, and enterprise technology organizations. Rather than debating feature sets, the panel focused on ownership, governance, and alignment with organizational risk tolerance.
A consistent theme emerged. Trust is not solely a function of accuracy. It depends on predictable behavior, clear escalation paths, and human involvement where judgment matters. Agentic systems must fit within existing accountability models, not bypass them. Human-in-the-loop design was framed not as a fallback, but as a core mechanism for maintaining control and confidence.
The conversation reflected a broader maturity in the ecosystem. The focus is shifting from proving that agentic systems can work to ensuring they work responsibly, consistently, and in service of real business outcomes.
5. Industry and people context still shape whether technology delivers value
Alongside deep technical discussions, two complementary themes continued to surface. Industry context and human capability remain decisive factors in whether technology succeeds.
After HITEC, Aquiva’s Nikolai Balba recapped insights on how hospitality organizations are navigating AI, data integration, cybersecurity, and personalization (check it out here). The emphasis was not on novelty, but on practical adoption. Unified data, ethical use of AI, and security awareness are becoming prerequisites for guest experience innovation, not differentiators.
That same grounding showed up in sessions focused on soft skills and stakeholder alignment, including reflections shared by Tatsiana Belavusava within the Salesforce community. These conversations highlighted a parallel reality. Technical excellence alone is not enough. Communication, empathy, and the ability to navigate stakeholder dynamics directly influence delivery outcomes, especially in partner-led ecosystems.
Taken together, these themes reinforced a simple truth. Technology does not operate in isolation. Industry constraints, organizational dynamics, and individual growth all shape whether platforms like Salesforce and Agentforce translate into meaningful results.
Closing reflection
Where we showed up
Beyond the sessions highlighted above, Aquiva team members participated in Salesforce and industry events throughout the year. Speaking, attending, and contributing across Dreamin’ conferences, World Tours, community groups, and industry forums.
These moments matter not because of their volume, but because they reflect ongoing engagement with the ecosystem. The conversations continue, and so does the learning.
And so, we want to say thank you to all our employees that took time away from their families this year to share, network, and learn to make us and the ecosystem better!
Author
Aquiva Labs
When you count on Salesforce, count on Aquiva.
