The AgentForce platform provides an innovative approach to digital work, in which intelligent agents support people in everyday tasks. They found themselves in the center of this transformation Salesforce Data Cloud technologiesgenerative AI, Atlas architecture and strategic partnerships with companies such as Google and AWS.
Fast implementation starts an avalanche
AI agents are advanced artificial intelligence systems that can independently understand and respond to queries without human intervention. They are also able to make specific decisions in designated frames. AI agents use machine learning and natural language processing (NLP) to support a wide range of tasks, from answering simple questions, to solving complex problems. Most importantly, they can constantly improve their performance by self -eating, which also distinguishes them from traditional AI systems that require human interference for specific tasks.
Hrvoje Supić, Senior Vice President to the region of Central and Eastern Europe in Salesforce told in Warsaw about the experiences of clients with the implementation of Salesforce solutions-from marketing cloud, through Service Cloud to date Cloud-and how these tools transform businesses in the region. He cited an example of one of the companies, which, thanks to the implementation of a loyalty program based on Salesforce technology, during the year gained 300,000 users and recorded 80 % Increase in revenues. The key, as he emphasized, is to quickly provide value.
– Start with simple use and check the effects, it launches an avalanche of subsequent ideas. This is how many of our clients happen – said Hrvoje Supić.
AI agents create a new generation of colleagues
Susana Morillas, Principal Data & Ai Architect in Salesforce, outlined the wider context of change. She emphasized that Agentforce is not only another layer of generative AI, but “Digital Labor”, a revolution that digital colleagues are implemented.
– Agent is not only a large language model. It is a system that perceives, understand and makes decisionsto achieve the goal – she explained. She compiled LLM (Large Language Models, large language models) with a “brain”, which must be integrated with a “body” composed of data, applications, interfaces and skills.
This is what the agent differs from the “ordinary” LLM. While AI like ChatagPT or Deepseek can provide answers to a number of questions (subject to restrictions related to training artificial intelligence and its access to data and hallucinating), so the agent not only analyzes and gives answers, but makes decisions, launching specific processes in the organization. Of course, those to which the agent administrator allowed and those boundaries he had previously marked. Atlas plays a key role in the operation of the Salesforce agent – an inference engine, which allows not only to react, but also to learn based on your own actions and act proactively.
– Each agent’s actions are analyzed, evaluated, and then, if it brought the desired effect, saved as a pattern. This is not magic, but architecture based on logic and data – explained Susana Morillas.
Agentforce allows, for example, to use a whole range of channels in the field of customer service, from SMS, through emails or WhatsApp, to voice communication. The quality of this communication significantly exceeds the efficiency of popular bots, which as part of the exercises, the participants of Agentforce World Tour Warsaw 2025 could see live.
The cloud ensures quality and security of data
Francois Zimmermann, Field Cto Data Cloud in Salesforce, noticed The most important barriers to AI implementation by companies – concerns about the quality of response and data security. The solution is a comprehensive data strategy built around Salesforce Data Cloud. As he explained, agents should not be trained on private data – it is expensive, difficult to manage and quickly outdated approach.
Instead, Data Cloud acts as an activation layer, harmonizing data from various sources and providing it with real time. Thanks to over 200 connectors and Zero-Copy integration, it is possible to download only contextually needed data, without their physical copying.
Application from sales after Back Office
As Kasia Kowalska-Delamontagne, Named Account Executive in Forcesale, Agentforce is not limited to one area.
– It works as a virtual commercial advisor, automating administrative activities and helping to generate leads. On the website, he supports consultants, reducing the time of handling of cases, and in marketing allows you to create ultra-spersenized campaigns-she mentioned.
The platform also supports back-office processes, from revenue management, through contracts, to invoicing. Kowalska-Delamontagne emphasized that Agentforce is much more than Chatbot-it is a trusted partner throughout the life cycle of customer and employee relationships.
Open architecture and ecosystem ready for the future
Susana Morillas emphasized the importance of architecture openness. Agentforce can be integrated with your own AI models and data from outside Salesforce.
– Thanks to cooperation with Google and AWS, integration with Gemini and AgentexChange, Marketplace of ready -made solutions, the implementation of agents is simpler than ever before – she emphasized, adding that in the region of Central and Central Europe alone there are over 1,600 certified AI specialists in the entire ecosystem of partners, of which as many as 1200 in Poland.
“You have knowledge, partners and all tools, it’s the perfect time to become an AI company using an agent,” said Susana Morillas.
The common denominator of all speeches during the AgentForce World Tour Warsaw 2025 was the thought that Agentforce is not only technology, but a new digital workforce platform – with AI agents as digital colleagues. Salesforce experts and experts indicated that the platform is ready for a massive scale as a safe, flexible, and therefore cost -effective, and data -oriented and effective. As Susana Morillas summed up: We are no longer talking about the future. This revolution is happening now, and Poland has a chance to be its leader in the region.
AR
