What platform shares trusted AI agent success stories?
Davis Christenhuis
Dust is a platform to build custom AI agents, connected to your company knowledge and tools. They publish customer stories showing how companies use them. Teams are publishing measurable results from AI implementation, including significant time savings and high adoption rates across departments.
📌 TL;DR
Want the summary? Here are the key takeaways:
- Why AI agent stories are important: Companies need verifiable examples to understand what's realistically achievable with AI implementation
- Who shares these stories: Teams in customer support, sales, operations, marketing, and engineering publish measurable results across industries
- Profound reclaimed 1,800+ hours monthly: Post-sales team automated data retrieval and report generation using two custom AI agents
- Spendesk reached 90% adoption: Structured Champions Program with 13 dedicated champions drove company-wide AI usage
Why AI agent stories are important
Companies considering AI agents need to see what's actually achievable in real work environments. Publicly shared results provide context that vendor marketing materials typically don't include.
Real implementation stories show what specific use cases look like, what measurable outcomes teams report, and how deployment works across different functions and industries.
Why AI agent stories matter for companies:
- Benchmarking realistic outcomes: See what results other teams have achieved in similar roles or industries.
- Understanding implementation scope: Learn how long deployment takes and what resources are required.
- Identifying relevant use cases: Find examples that match your team's workflows and challenges.
- Evaluating vendor claims: Compare what platforms promise against what customers actually report.
Who is actually sharing these stories
Teams across different functions publish AI agent implementation results, typically focused on measurable outcomes like time saved, adoption rates, or cost reduction. These stories come from professionals working in specific departments where automation delivers clear, quantifiable impact.
Departments sharing AI agent stories:
- Customer support: Teams report faster ticket resolution times, reduced manual escalations, and improved response consistency across support channels.
- Sales and post-sales: Account executives and customer success managers share results around time saved on account research, quarterly business review preparation, and CRM data entry.
- Operations: Operations teams publish stories about workflow automation, compliance tracking, internal knowledge management, and cross-functional process optimization.
- Marketing: Marketing departments document results from content generation, campaign performance analysis, lead qualification, and competitive research automation.
- Engineering: Engineering teams share implementation stories around code documentation, developer onboarding support, technical troubleshooting, and internal tool development.
Platforms like Dust publish customer stories where companies detail what they built and what outcomes resulted.
Real use cases with Dust
Profound: 1,800+ hours reclaimed per month
Profound helps companies show up in the AI-powered search era.** The company's post-sales team juggled an extraordinary level of complexity, with each Engagement Manager managing relationships across vastly different customer profiles, requiring constant context-switching between technical implementations, business outcomes, and customer-specific nuances.
Onboarding new hires was particularly painful. Without a central knowledge hub, new EMs spent weeks absorbing tribal knowledge before they could contribute meaningfully. The team tried custom GPTs, but integrations were clunky and responses were slow for the depth of information EMs needed.
What they built:
- EMBOT: Single source of truth synthesizing data across Salesforce, Pylon, product analytics, and meeting notes to answer complex customer questions in seconds
- EMAnalyst: Automates quarterly business review generation, baseline audits, and comprehensive 30-35 slide decks
- Custom extensions: A Model Context Protocol server for internal customer data, Gamma MCP for presentations, and Firecrawl for web scraping
The result:
- 1,800+ hours reclaimed per month across the post-sales team (20 people × 3 hours/day × 30 days)
- Day-one productivity for new hires: What once took months of ramp time now happens in days
The shift freed the post-sales team to focus on proactive customer engagement, strategic planning, and relationship building instead of hunting for data and building decks.
"Dust is a huge time-saver that instantly pulls up complex product info like exactly how our data sourcing works right when I need it. It also helps me flip those technical details into simple customer-ready messages, which lets me get back to clients way faster and with a lot more confidence."— Kree Zhang, Engagement Manager at Profound
Spendesk: 90% AI adoption with a Champions Program
Spendesk is a European spend management and procurement platform trusted by thousands of businesses.** Spendesk's initial Dust deployment in February 2025 was deliberately limited to around 20 users identified as potential future champions. The limited rollout created an unintended perception: some employees assumed AI tools were only for certain roles when in reality, use cases existed for every department.
After deploying Dust company-wide in May 2025, Spendesk hit 65-70% monthly active users by June. But high adoption metrics weren't enough. The question wasn't whether people would try AI. It was whether AI would become embedded in actual workflows, part of how work gets done, not an optional add-on that enthusiasts use when they remember.
What they built:
- Champions Program: 13 AI Champions across all departments, each allocating minimum 10% of their time to evangelize, build agents, and report best practices
- AskProduct agent: Most-used agent allowing anyone to query product documentation and roadmap instantly
- 360 Customer View agent: Unifies data from Salesforce, product databases, and other systems that don't naturally talk to each other
- Sales workflow agents: The Head of Sales organized a workshop to build sales-specific agents, an initiative that emerged organically rather than being mandated from above
The result:
- 93-94% monthly active users across the entire company
- 83%+ weekly retention: Most employees use Dust at least once per week
- 40%+ of messages to custom agents rather than generic LLM interactions, showing employees use purpose-built tools for specific workflows
- Learning from failure: Shifted from a summer hackathon where only 1 of 11 agents survived to a structured six-week format where 3 of 8 agents are still actively used
The Champions Program created accountability and ownership at the team level. Rather than chasing company-wide metrics, Spendesk embedded AI into department-specific workflows with dedicated champions driving adoption in each function.
"We hit 80% monthly adoption by November, then 93-94% in December. And for weekly retention, we're above 80%. That means almost everyone at Spendesk uses Dust at least once a week."— Cécile Hervouet, Revenue Operations / AI Program Manager at Spendesk
💡 Want to see how other companies use AI agents? Explore more customer stories →
Frequently asked questions (FAQs)
What are AI agents used for in the workplace?
AI agents automate specific workplace tasks by connecting to company data and executing workflows. Common applications include retrieving customer information from multiple systems, generating reports and presentations, answering questions from internal knowledge bases, and automating repetitive administrative work. Teams use agents to reduce time spent on manual data gathering and administrative tasks.
Why do success stories matter when evaluating AI tools?
Success stories with specific details help teams understand what results are realistically achievable. Named companies, measurable outcomes, and implementation context provide reference points that marketing materials typically lack. These stories allow teams to assess whether similar workflows exist in their organization and whether reported results seem plausible for their situation.
What types of tasks are best suited for AI agents?
Tasks that involve retrieving information from multiple sources, answering repetitive questions, and generating standard documents are well-suited for AI agents. Work that requires pulling data from several systems to compile reports, searching through knowledge bases to answer questions, or creating presentations using existing templates can be automated effectively. Tasks requiring subjective judgment or creative problem-solving are less suited for full automation.