Paul Nashawaty, Practice Lead & Principal Analyst, AppDev Practice, theCUBE Research
Dreamforce has always been a bellwether for enterprise innovation, particularly in the Salesforce ecosystem. In 2025, the event reinforced its position as a platform where developers, partners, and enterprises come together to explore how AI, automation, and data are changing application development. As someone who has studied developer trends across industries for years, I found this year’s Dreamforce especially significant for its implications on AI-native development, low-code adoption, and the evolving role of developers.
The Agentic Enterprise and Redefining Application Development
One of the most important themes at Dreamforce 2025 was the notion of the Agentic Enterprise. Salesforce is now emphasizing an environment where intelligent agents, not just humans, drive business processes, orchestrate workflows, and automate decisions in real time. For developers, this shift is both exciting and challenging: it requires blending traditional coding skills with AI orchestration, data integration, and workflow design.
During the Developer Keynote, Salesforce highlighted new capabilities in Einstein, Data Cloud, Agentforce, and the Salesforce Platform SDK. Generative AI tools and automation frameworks are no longer optional, rather, they are becoming central to enterprise application design. Developers who can master these tools will have a greater impact in shaping how organizations operate in a more automated and data-driven world.
Heroku Bridges Open Development and Enterprise AI
At Dreamforce 2025, I have to mention Heroku, Salesforce’s open and extensible PaaS platform. Once primarily a playground for startups, Heroku now plays a strategic role for enterprise developers. It allows teams to deploy AI-enabled applications, integrate with Salesforce data, and extend Agentforce workflows using familiar languages and frameworks.
From my perspective and research, Heroku represents the ideal bridge between enterprise-grade governance and open, flexible development environments. For developers, this means you can leverage modern microservices architectures, integrate external APIs, and experiment with AI models while maintaining the security, scalability, and compliance required in large organizations.
Reinforcing Heroku’s Role in an AI/Integration-First Stack
One of the threads in recent MuleSoft announcements is the increasingly tight coupling of API/integration infrastructure with AI agents and cross-system automation. For instance, MuleSoft has pushed forward features enabling AI agents to access and act on data across CRM and external systems, and a more seamless integration between MuleSoft, Flow, and Heroku APIs.
For Heroku, this means its PaaS becomes more strategically central. MuleSoft’s support for deploying Anypoint Flex Gateway on Heroku (especially within private spaces) allows developer teams to expose and manage APIs hosted on Heroku with full governance, authentication, and traffic control, all via MuleSoft’s control plane. In effect, Heroku isn’t just “some app host” in the Salesforce ecosystem; it becomes an integral part of the API fabric, with the same governance, security, and observability that MuleSoft users expect.
Further, in the 2025 Heroku announcements, Heroku is recasting itself as an AI PaaS, not just a platform for web apps, but a platform that provides managed inference, agents, and deep integration with data services. That paves the way for developer teams to build intelligent, agent-driven experiences natively on Heroku, while leveraging MuleSoft’s APIs to bridge to enterprise systems, data clouds, and event streams.
In short: for Heroku, the Dreamforce MuleSoft news elevates its strategic role. Rather than being peripheral, it becomes a first-class citizen in the integration/agent architecture. That gives Heroku users stronger footing to build enterprise-grade, AI-powered applications without stitching together disparate components.
Impacts on Developer Experience, Productivity, and Ecosystem
From the perspective of developers and engineering teams, the MuleSoft announcements introduce several significant implications. They promise to reduce integration friction through unified governance and low-code tools while accelerating the development of AI- and agent-driven applications. At the same time, the deeper convergence of MuleSoft, Salesforce, and Heroku offers greater efficiency but raises new considerations around ecosystem lock-in, architectural complexity, and skill evolution.
Developers will need to adapt to more sophisticated integration models, emphasizing API design, event-driven patterns, and intelligent orchestration to fully leverage the new capabilities while maintaining agility and control.
- Lower friction in integration & governance
One perennial pain point is that building integrations, especially secure, well governed APIs — is tedious and error-prone. With MuleSoft’s push to embed low-code integration features into Flow, and tools like the Anypoint Flex Gateway for Heroku, developer teams can leverage a unified control plane for APIs across Salesforce, Heroku, and beyond, thereby reducing the burden of “glue code.” - Faster time to value for AI/agent features
Because MuleSoft is heavily orienting toward enabling AI agents (e.g., letting them call APIs, act across systems, and manage knowledge/data sources), developers get a more consistent path to embedding intelligence. Rather than patching together disparate services, the stack becomes more opinionated and integrated. - Potential ecosystem lock-in pressures
As MuleSoft, Salesforce, and Heroku integrate more deeply, the appeal of this unified stack grows. But this also raises questions for developers and organizations about flexibility: to what extent might they be locked into the Salesforce/MuleSoft/Heroku ecosystem versus alternative integration or service platforms? That tradeoff becomes more salient as the features deepen. - Opportunities for new tooling, extensions, and third-party innovations
Where there is rich API management, agent orchestration, and integration tooling, there’s fertile ground for third parties to build connectors, agent skills, monitoring, and extensions. Developers and ISVs can build on top of this expanded stack — e.g. new “actions” or skills for agents that call Heroku-hosted services, or monitoring tools for API traffic across systems. - Upskilling and evolving best practices
As the architecture shifts from “app + custom integration code” to “agentic architecture + governed APIs + integrated data fabric,” teams may need to evolve their skill sets. API design, security, event-driven architecture, AI prompting and agent orchestration become more central. The announcements implicitly encourage developers to level up on those domains.
Some Risks, Challenges, and Questions to Watch
As integration between Heroku, MuleSoft, Salesforce Flows, and AI agents deepens, developers face a new set of operational and architectural challenges. Performance and scaling complexities become more pronounced as additional routing, governance, and orchestration layers introduce potential latency, concurrency, and observability concerns. Maintaining consistent security, schema contracts, versioning, and error handling across these interconnected systems is equally vital to avoid weak links between services.
Existing Heroku applications may also require refactoring or migration to align with new gateways, agent hooks, and unified API catalogs, demanding a smooth transition to preserve developer confidence. Finally, while MuleSoft’s abstractions and low-code capabilities simplify integration, experienced teams will still seek granular control over routing, retry logic, and error policies, making it essential to balance ease of use with the flexibility needed for complex, enterprise-grade environments.
- Performance and scaling complexity
With more routing, governance, and cross-system orchestration, latency, concurrency, and observability challenges become more acute. Ensuring that APIs, agents, and data flows remain performant under load will be critical. - Consistency across layers
When integration spans Heroku, MuleSoft, Salesforce flows, and AI agents, ensuring that security, schema contracts, versioning, error handling, and observability are consistent is nontrivial. Developers must ensure that the “seams” between systems don’t become weak links. - Migration burden for existing apps
Teams that already run significant logic or APIs on Heroku may need to refactor or retrofit to adopt the new gateways, agent hooks, or unified API cataloging. That migration pathway must be smooth to maintain developer confidence. - Balancing abstraction with control
While letting MuleSoft provide more abstraction (e.g. low-code integration or AI-driven generation) is powerful, expert teams will still demand fine-grained control (e.g. custom routing, retry logic, error policies). Striking that balance is tricky.
Overall, the MuleSoft announcements at Dreamforce signal a tighter, more integrated future between Heroku and the broader Salesforce ecosystem, one oriented around API governance, AI agent orchestration, and seamless data flows. For developers, this promises lower friction, faster delivery of intelligent features, and new opportunities, but also invites new architectural discipline, awareness of tradeoffs, and careful attention to performance, security, and maintainability.
Learning, Experimenting, and Scaling
Dreamforce 2025 also reinforced the importance of hands-on learning. The Trailblazer Forest offered workshops, labs, and coding sessions where developers could explore new tools and approaches in real time. From generative AI integrations to Lightning Web Components and Flow Orchestration, the sessions were designed to accelerate developer skill growth, reduce barriers to innovation, and promote adoption of best practices across the ecosystem.
Our research at theCUBE has consistently shown that when developers have access to both low-code and pro-code tools, deployment frequency, innovation, and business impact increase substantially. Dreamforce 2025 demonstrated that Salesforce is making this dual approach a central part of the developer experience.
theCUBE AppDev Insights – Developer Trends and Market Signals
From an analytical perspective, Dreamforce 2025 highlighted several trends shaping the future of enterprise application development. Low-code and no-code adoption is accelerating, with 70% of enterprises we surveyed now empowering business users to build applications, while developers focus on integrating, securing, and scaling these solutions.
At the same time, AI has become table stakes; intelligent agents, predictive workflows, and automation frameworks are now expected in enterprise applications. This requires developers to quickly adapt and incorporate these capabilities. Finally, ecosystem influence remains a key differentiator: Dreamforce offers unmatched access to Salesforce product teams, partners, and peers, which enables developers to shape product direction and bring innovative solutions to market faster. To summarize, several trends stood out this year:
- Low-code/no-code adoption is accelerating – 70% of enterprises we surveyed are now using low-code platforms to empower business users. Developers are shifting toward integrating, securing, and scaling these applications.
- AI is table stakes – Intelligent agents, predictive workflows, and automation frameworks are now expected in enterprise applications. Developers need to adapt quickly to incorporate these tools.
- Ecosystem influence matters – Dreamforce provides unparalleled access to Salesforce product teams, partners, and peers, allowing developers to influence product direction and bring innovative solutions to market faster.
Comparing Dreamforce to Other Developer Conferences
While AWS re:Invent, Microsoft Build/Ignite, and Google I/O all provide excellent technical depth, Dreamforce stands out for its enterprise focus, ecosystem alignment, and community-driven innovation. Unlike AWS or Google, which often emphasize infrastructure or open-source tooling, Dreamforce equips developers to solve business problems, drive automation, and innovate in customer engagement workflows.
Heroku, Data Cloud, and Agentforce all enable developers to bridge these worlds, combining flexible application development with enterprise-grade AI and security. In short, Dreamforce delivers a uniquely holistic developer experience where technology, business impact, and community converge.
Analyst Take – Dreamforce as a Launchpad for Developers
For developers, Dreamforce 2025 was a reminder that their role is evolving. No longer just coders, developers are becoming architects of autonomous, intelligent systems. By embracing AI, low-code, and hybrid development environments, they can shape how enterprises operate, innovate, and deliver value.
Five Next Steps for Developers to Take After Dreamforce 2025
- Experiment with AI tools – Try Einstein, Agentforce, and generative AI to automate workflows.
- Use Heroku for flexible apps – Deploy microservices and AI-enabled applications quickly and securely.
- Mix low-code and pro-code – Let citizen developers build simple apps, while pro developers handle integrations and scaling.
- Engage with the Salesforce community – Join forums, workshops, and partner events to share ideas and learn best practices.
- Focus on business impact – Prioritize projects that improve customer experience, efficiency, or revenue.
From my perspective, and aligning with my research, Dreamforce acted as more than just an event. It was a strategic touchpoint for developer empowerment. It demonstrates where enterprise software is heading and gives developers the tools, insights, and community support needed to lead in the era of the Agentic Enterprise.
Dreamforce 2025 Developer Impact and the Rise of the Agentic Enterprise
The AI Infrastructure Paradox: When Buying GPUs Is Just the Beginning
Databricks’ Mooncake Move and OpenAI Pact
Developers Face Growing Consumer Distrust Over AI Code Security
Google Cloud Advances AI and Sovereign Infrastructure Across Public Sector
