The creation of Agentforce was inspired by a real-world problem faced by field sales representatives, their managers, and admin teams who often need quick access to critical customer information—such as order history, product details, discounts, pricing of products, new launches, and current offers—while on the move or during client meetings. The goal was to eliminate the friction of navigating through multiple Salesforce screens and instead deliver relevant data seamlessly within Slack, a tool already embedded in their daily workflow.
We started by building on the existing Agentforce framework, identifying sales enablement as our core use case. Our team of four went through multiple internal discussions to refine the idea and finalize the problem statement.
During the development phase, we encountered several challenges, especially around Slack configuration and the limited feature availability of Agentforce at that time. Despite these roadblocks, we proactively worked with Salesforce support, raised tickets, and iteratively fine-tuned our solution.
We also focused heavily on optimizing the responses from Agentforce by improving how it handled both structured and unstructured data. Through targeted data modeling, careful prompt adjustments, and flow optimization, we ensured the system delivered precise, actionable insights.
What started as a vision to empower frontline teams evolved into a highly functional AI-driven assistant—and ultimately, a winning solution recognized at the Salesforce TDX 2025 Hackathon for Best Use of Slack.
The Dream Team That Brought Agentforce to Life
The success of Agentforce is a direct result of our tightly-knit, cross-functional team. Each member brought a unique strength:
Our approach to integrating Salesforce with Slack was both architectural and experience-focused. We aimed to create a seamless, context-aware interaction layer where users could query Salesforce data without ever leaving Slack. The foundation of our integration was built on Salesforce Slack SDK, Data Cloud, Apex, Flows, and VS Code for development.
Early in the process, we faced several technical hurdles—particularly in configuring Slack to work effectively with Salesforce and adapting to the evolving Agentforce capabilities. We worked around these challenges by engaging with Salesforce support, understanding the system limitations, and developing custom logic to bridge functionality gaps.
One of the key aspects of our solution was the refinement of the bot’s ability to interpret user queries accurately. This involved enhancing the handling of both structured Salesforce data and unstructured user inputs. We designed flows that ensured real-time, relevant responses—aligned with business needs and sales workflows. Our focus on user-centric design, robust data handling, and continuous iteration allowed us to deliver a smooth integration experience.
Our collaboration, adaptability, and shared sense of ownership were key to overcoming obstacles and delivering a winning solution.
Code, Chaos & Comebacks: The Real Hackathon Story
Our journey through the hackathon was filled with both technical challenges and breakthrough moments. It started on a rough note—the first two hours were lost due to internet connectivity issues, which significantly delayed our initial setup and development kick-off. This early setback added time pressure and forced us to work even more efficiently as the event progressed.
Once we resumed work, we moved forward quickly with building the Slack-based assistant using Agentforce, Salesforce Data Cloud, Apex, and Flows. However, just when we thought we had implemented the core functionality, we ran into a major issue—Agentforce wasn’t giving the correct responses in the chat window, even though our debugging logs showed the right results.
This inconsistency was puzzling and consumed a large chunk of our development time. We initially suspected a bug in our logic or data handling, but after exhaustive debugging, we reached out for help.
Several Salesforce staff members on-site stepped in—three to four experts examined our code, flows, and configurations. Eventually, one of them identified the root cause: a specific checkbox setting within Agentforce was modifying the final output in the chat interface, even though the backend logic was functioning correctly.
This revelation was a turning point. Once we corrected the configuration, the assistant started responding accurately, as intended. That one checkbox had caused a subtle yet critical difference between what we expected and what users saw.
Despite the initial delays and unforeseen issues, the team stayed composed, collaborative, and solution-focused. Dividing responsibilities and keeping constant communication allowed us to keep making progress. These challenges, although frustrating in the moment, ultimately made the win more meaningful.
Iterate. Test. Win. The Agentforce Build Journey
The build process was both structured and adaptive. We started with a series of rapid brainstorming sessions to finalize a high-impact use case—helping field sales reps and managers access real-time Salesforce data directly within Slack. Once aligned, we divided the responsibilities across four core areas: Salesforce data modeling, Apex and Flows, Slack integration, and user experience design.
We began building on top of the Agentforce framework using VS Code and integrated it with Slack. Early test responses were off-mark, especially for queries involving product or offer-related data. We spent considerable time refining both the structured and unstructured data, optimizing prompt design, and validating how Agentforce interpreted the context.
The process involved continuous iterations. We followed a build-test-refine loop, verifying every feature in real-time Slack conversations. When things didn’t work as expected, we traced logs, adjusted data mappings, and tested edge cases.
After resolving the configuration issue with help from Salesforce staff, we spent the final hours polishing the experience—adding fallback messages, adjusting UI formatting in Slack, and making responses more user-friendly and contextual.
Secure. Smart. Seamless. What Makes Agentforce Stand Out
Agentforce is uniquely engineered for Salesforce environments, combining real-time AI, Slack-first integration, and robust enterprise-grade security. Its seamless connection to Salesforce Clouds and Slack allows users to interact with data through natural language while ensuring all activity respects enterprise governance.
What truly sets Agentforce apart is Salesforce’s Trust Layer—which ensures sensitive data is masked and never stored—making it ideal for industries with strict compliance requirements.
Additionally, Agentforce leverages Atlas, Salesforce’s reasoning engine, to understand user context and maintain conversation flow without persisting customer data. This enables accurate, dynamic responses tailored to the user’s role, intent, and real-time Salesforce context—whether they’re in sales, service, or operations. The result is a powerful blend of automation, security, and usability that few platforms in the market can match.
Work Smarter, Serve Faster: Agentforce in Action
Agentforce empowers users to work faster and smarter by bringing contextual AI directly into Slack, reducing tool-switching and manual searches. Sales reps, managers, and service teams can instantly access account details and order histories or create records—guided by AI that understands their goals and follows enterprise policies. Its automation not only accelerates decision-making but also ensures that processes are compliant and secure by design.
From a customer experience standpoint, Agentforce delivers AI-assisted, human-like support with real-time insights—all while keeping customer data protected. The use of masked data, non-persistent memory, and secure access ensures trust is never compromised.
By enabling faster, more personalized responses, Agentforce helps businesses strengthen relationships and boost satisfaction without sacrificing security or control.
Vision for the Future Next Stop: The AI-Driven Enterprise Workspace
In the next 2–3 years, Agentforce is poised to become the default intelligent interface across Salesforce ecosystems, moving beyond basic task automation to fully contextual, real-time enterprise assistance.
With continued advancements in Salesforce’s Trust Layer, Atlas reasoning engine, and generative AI, Agentforce will evolve into a more proactive solution—anticipating user needs, suggesting next-best actions, and automating multi-step workflows across clouds and third-party systems.
We expect it to integrate even more deeply with Data Cloud and Flow Orchestration, enabling cross-functional use cases that blend sales, service, marketing, and operations.
As AI maturity grows within Salesforce, Agentforce will likely become a core enabler of composable enterprise experiences, driving intelligent automation while ensuring data security, compliance, and user personalization at scale.
Beyond Retail: Agentforce’s Industry-Agnostic Superpowers
Beyond retail, Agentforce has immense potential across healthcare, financial services, manufacturing, and public sector industries—anywhere data-driven decision-making and rapid response are critical.
In healthcare, for instance, it could support clinicians and care coordinators with real-time access to patient data while preserving privacy through Salesforce’s masking and trust controls. In finance, it can assist advisors with secure client insights, portfolio updates, and compliance workflows.
The combination of contextual AI, Slack-native interaction, and enterprise-grade security makes Agentforce a cross-industry enabler of operational efficiency and intelligent service delivery.
As industries continue to adopt AI and unified data strategies, Agentforce will help transform how frontline teams interact with systems—bringing speed, clarity, and intelligence into everyday workflows.
Fixing the Fracture: Solving Real Business Pains
One of the biggest challenges businesses face today is navigating fragmented systems, slow access to insights, and rising data security concerns. At the same time, there’s a clear industry trend toward AI-driven automation, contextual service, and connected digital experiences.
Agentforce is designed to address these head-on—offering secure, real-time access to enterprise data through conversational AI within Slack, without exposing sensitive information.
With features like the Salesforce Trust Layer for data masking, Atlas reasoning engine for context retention, and native integration across Salesforce Clouds, Agentforce bridges the gap between speed and compliance. It tackles the demand for faster decision-making, cross-functional visibility, and intelligent workflows—while ensuring enterprise-grade governance and control.
Stay Ahead or Stay Behind: Agility Through Agentforce
Agentforce empowers businesses to stay ahead by making real-time, AI-assisted decision-making accessible to every role—from sales reps to service agents and ops teams. Instead of relying on rigid processes or switching between tools, teams can engage with enterprise data directly in Slack—speeding up execution and improving collaboration across departments.
Built to Scale: What Agentforce Brings to the Table
Salesforce’s Agentforce provides businesses with autonomous AI agents capable of executing tasks across various functions, including sales, service, marketing, and commerce.
These agents are designed to operate 24/7, integrating seamlessly with the Salesforce ecosystem to enhance productivity and customer engagement. By automating routine processes and facilitating real-time decision-making, Agentforce aligns with businesses’ needs for efficiency and scalability in an increasingly digital landscape.
The Magic Behind the Messages: Agentforce Highlights
Key features of Agentforce that benefit customers include the Atlas Reasoning Engine, which enables agents to understand context and make informed decisions autonomously. The platform also emphasizes data security through Salesforce’s Trust Layer, ensuring sensitive information is handled appropriately. Additionally, Agentforce’s integration with Slack allows for seamless collaboration and communication, enhancing the overall user experience.
See the Full Press Release on our TDX Hackathon Success Here.
How do you feel about the team’s win at the TDX Hackathon?
Winning the TDX Hackathon was a proud and validating moment for our team. One key takeaway is that agentic systems are clearly the future—we’re moving away from clicking through screen-by-screen interfaces to interacting with intelligent agents that deliver precise, context-aware results in seconds.
The experience showed us that well-structured APIs, no-code flows, and clearly defined rules (both structured and unstructured) will be critical for AI systems to interpret and act on business logic effectively. We also saw how critical it is to fine-tune backend configurations to make AI outputs meaningful and accurate.
Watch our Best Use of Slack demo from the Agentforce team at the TDX Hackathon to see how intelligent automation transforms sales and service workflows in action.
AI Builders, Take Note: Lessons from the Frontline
Start with solving real problems—AI is only as good as the clarity of the system it works with. Today, AI might help with informational queries, but soon, chat will become the new UI, and English will be the new programming language. Developers will need to think beyond code and act more like consultants—deeply understanding business needs and translating them into intelligent, adaptable systems. Stay curious, stay grounded in user needs, and build with scalability and context in mind.
Ready for Real AI ROI? Let’s Build the Next Big Thing
Our next priority is to equip every team member with the skills to effectively work with AI and Agentforce, enabling us to deliver tailored solutions that solve real customer challenges. For businesses interested in Agentforce, the key first step is to ensure their CRM systems are well integrated and their data flows seamlessly.
Once the foundation is solid, building AI-powered experiences on top becomes far more effective—allowing Agentforce to operate smoothly and deliver exceptional value. We’re ready to partner with organizations at any stage to help unlock this potential.
