Our CTO, Praveen Ravela, recently attended Trailblazer DX (TDX) 2026, spending two days in deep-dive conversations with Salesforce product managers, solution engineers, and technical architects. Below, we explore the core themes of the event, what they mean for CX leaders, and how organizations can bridge the gap between visionary AI roadmaps and enterprise-grade deployment.
1. Building Trust into the Platform Core
Salesforce is doubling down on the “Trust Layer” by adding native observability and automated testing to Agentforce. A key milestone to watch is June 2026, when bot-based testing for Agentforce Voice agents is expected to launch.
One powerful takeaway for improving AI consistency was the shift toward Vibe Engineering. Rather than relying on open-ended generative behavior, Salesforce is encouraging architects to feed structured scripts and defined outcomes directly into the AI engine. This hybrid approach ensures AI agents operate within a predictable, high-stakes framework while maintaining the flexibility of generative intelligence.
2. The Evolution of Multi-Agent Orchestration
Multi-agent A2A (agent-to-agent) orchestration is slated for June 2026. This architecture introduces the “Super Agent” model, a coordinator that manages Sub-Agents and external systems. This moves Agentforce from isolated deployments toward truly coordinated workflows that span diverse data sources.
Note: Native Salesforce A2A orchestration and MuleSoft-led orchestration remain distinct tracks, allowing enterprises to choose the path that best fits their existing integration maturity.
3. Navigating the Hybrid CCaaS Ecosystem
One of the most candid themes at TDX was the complexity of the modern contact center. While Salesforce is innovating rapidly with Agentforce Contact Center, the reality for most enterprises involves a multi-vendor environment.
Current market challenges such, as SIP trunk provisioning and the alignment of CCaaS partner bots with Agentforce Voice, require a thoughtful architectural approach. For many, the goal isn’t a platform swap, but rather a “co-existence” strategy where calls can transition between Agentforce AI and human agents on existing CCaaS platforms without losing context.
4. Slack’s Strategic Internal Role
The vision for Slack remains focused: it is the premier internal workforce collaboration hub. With Ambient Mode, Slack can now “listen” to live conversations for transcription and internal agentic workflows. While external, customer-facing voice remains off the roadmap for now, there is growing interest in how Salesforce Voice experiences can be surfaced within Slack to empower internal teams.
5. Headless 360: Salesforce as Infrastructure
Perhaps the most significant shift is Headless 360. Salesforce is positioning itself as a robust backend layer, providing the APIs, data, and orchestration, without requiring a specific UI. This “Salesforce as Infrastructure” model allows experiences to be rendered anywhere, from mobile apps and WhatsApp to custom interfaces. It supports a “Build Once, Use Everywhere” philosophy that aligns perfectly with modern enterprise architecture.
What to Watch After TDX 2026 (The TL; DR)
Much of what was discussed at TDX 2026 is still in motion. June 2026 is the key near-term date for multi-agent A2A orchestration and automated Agentforce Voice testing. For teams currently evaluating or piloting Agentforce, those releases directly address the predictability and oversight gaps that have made adoption difficult.
Headless 360 and the full Agentforce Experience Layer are earlier in development. Expect continued announcements, but plan for a longer timeline before broad enterprise availability.
For contact center leaders, it comes down to a few specific planning areas:
- CCaaS-to-Agentforce integration complexity remains unresolved; architecture decisions made now will determine how much friction you carry forward
- Agentforce Contact Center licensing costs require careful modeling, particularly for back-office use cases
- Slack belongs in internal workflows; customer-facing voice and interaction use cases are not yet on the roadmap
- Observability and testing tools are arriving in June. You should start designing your oversight processes before that lands
What This Means for Your Organization
TDX 2026 confirmed that AI in enterprise has shifted from the question of whether to deploy AI to how to operate it consistently, with visibility and control over time.
1. Training inputs determine Agentforce output quality.
Feeding structured scripts and defined outcomes into AI systems was highlighted as a key method for achieving reliable results. This applies broadly, not just within Salesforce. Organizations that invest in structured training data early will see more predictable AI behavior.
2. Salesforce AI risk lives at the integration layer
Every AI interaction moves data: through transcription services, APIs, CRM systems, and storage layers. These are the points where performance varies, and exposure accumulates. Visibility into the integration layer is the foundation of any credible AI governance posture. Read more on data ownership in your contact center here.
3. Most enterprises will run multiple platforms for some time.
Salesforce is building toward a unified AI layer. Existing CCaaS investments are not going away on that timeline. The organizations that fare best will be the ones that plan for that reality now, rather than waiting for the platform to catch up.
Our Perspective on TDX 2026
TDX 2026 reinforced what we see in practice with clients every day. AI needs to be observable. It needs to be testable. And trust is built across the full ecosystem, not within any single platform. The candid conversations we had at TDX, particularly around CCaaS friction and Agentforce Contact Center gaps, confirmed that the challenges our clients are navigating are real and recognized by Salesforce itself.
AMC Technology will continue to focus on keeping data within enterprise control, securing integration points, training AI with structured inputs, and supporting flexibility across Salesforce and existing contact center environments. That work is underway now for customers who cannot wait for the platform to resolve these questions on its own timeline.
Salesforce is making serious investments in enterprise AI trust. The direction is clear. The gaps are also clear. Observability, CCaaS integration complexity, licensing structure, and architectural maturity are all areas that require attention beyond what the platform provides today.
The organizations that will benefit most from where Salesforce is headed are the ones that build the right foundation now: structured training, controlled integrations, and architecture that spans existing systems and new capabilities.
If those questions are already on your radar, we are ready to help you work through them.
