Alan’s PMM team transforms sales conversations into intelligence with AI agents
About Alan
Alan is a leading digital health partner transforming healthcare by combining advanced technology with a human touch to make health journeys seamless and accessible. Founded in 2016 by Jean-Charles Samuelian and Charles Gorintin, Alan launched as the first new health insurer in France since 1986 and now operates in France, Spain, Belgium, and Canada. Backed by over €275 million from top investors, Alan today supports more than 720,000 members across Europe and beyond, making quality healthcare simpler and more personal.
Challenge: scaling adoption of the sales narrative and monitoring impact
Alan's Product Marketing team faced a significant challenge in monitoring sales narrative adoption across an increasing volume of sales conversations. With teams operating across multiple countries, they needed to ensure consistent message delivery and identify gaps in narrative execution across their expanding sales organization.
The manual process was overwhelming - three Product Marketing Managers (PMMs) spent 2-3 hours each per week reviewing just a sample of calls, leaving most conversations unanalyzed and narrative adoption insights incomplete.
Solution: automated AI-powered conversation analysis
Alan developed a comprehensive system combining their automation platform with specialized Dust AI agents to analyze sales call transcripts from their Modjo platform. The solution features a sophisticated workflow:
- Automated data extraction from Metabase
Using ActivePieces as their automation platform, the workflow automatically retrieves all Modjo transcripts from the internal database. This happens on a daily basis. The current scope focuses specifically on discovery calls, where narrative adoption is most critical for setting the foundation of customer relationships.
Step 1 to 5 in the visual.
- Individual transcript processing
Each transcript is extracted from the database into a separate file for individual analysis. This approach significantly improves analysis quality by allowing the AI agent to focus on scoring a single transcript at a time, avoiding hallucinations and mistakes that can occur when processing multiple conversations simultaneously.
Step 6 in the visual.
- Country-specific AI analysis
Individual files are analyzed by dedicated Dust AI agents. Alan deployed separate agents for each country to account for local narrative specificities and avoid cross-market confusion, ensuring accurate evaluation of region-specific messaging.
Step 7 in the visual.
Agent specificities:
- Country-specific AI agents for local market understanding
- Customizable extraction parameters for different insight types
- Structured scoring system for narrative adoption across key narrative aspects
- Automated JSON output integration with reporting systems
Extract of a prompt of one of these Dust agents:
- Evaluation framework
After the analysis by the AI agent, a JSON output with the scoring is sent back to the reporting sheet, where it’s ready for analysis. This structured output ensures a robust and replicable approach.
Step 8-9 in the visual.
Results: comprehensive insights with 80% less effort
The implementation transformed Alan's ability to monitor and improve sales conversations, with Dust agents augmenting the PMM team by automating the surfacing of hard-to-parse information, freeing them to focus on strategic analysis rather than manual data extraction.
“Before Dust, our PMM team spent 6-9 hours weekly analyzing just a sample of sales calls manually. Now, our custom AI agents automatically analyze ALL discovery calls, scoring them against our 5-block narrative framework with country-specific precision. We've gone from subjective, sample-based insights to comprehensive, data-driven intelligence delivered weekly to sales leads. Dust didn't just automate our process - it transformed us from reactive analysts to proactive strategic partners. We reclaimed 80% of our analysis time while delivering 10x the insights." - Evelien Hesters, Product Marketing Manager, Alan
The Dust agents transformed Alan's ability to monitor and improve sales conversations:
Operational efficiency
- Reduced analysis time from 2-3 hours/week/PMM to total automation
- Saved 18+ hours monthly on manual review across 3PMMs
- Automated weekly reports to stakeholders
Enhanced visibility
- Analysis of all discovery calls instead of just a sample
- Real-time tracking of narrative adoption across markets
- Consistent evaluation methodology across countries
Actionable outputs
- Weekly insights delivered directly to sales leads
- Global narrative adoption reports for strategic oversight
- Data-driven coaching recommendations for individual reps
“It's not just about scaling our analysis—it's about turning every customer conversation into strategic intelligence that was previously impossible to capture,” - Evelien Hesters, Product Marketing Manager, Alan
👉 Focus remains: qualitative insights - sample based
👉 Turbocharge our PMM power with 3 AI agents: get a full quantitative analysis
Beyond the PMM application: unlocking endless new use cases for Alan
A primary use case for this system has been monitoring and driving adoption of PMM narratives by the Product Marketing team. With this foundation in place, Alan can rapidly deploy new AI agents far beyond narrative monitoring - the team can assess prospect reactions and sentiment, track feature adoption patterns, and even monitor regulatory compliance across different markets. Every customer conversation turns into a strategic asset that drives product development, competitive positioning, and market expansion decisions.
"The same workflow can be adapted to extract any type of information we need from our customer conversations, making it an invaluable tool for product, marketing, and sales teams alike." - Evelien Hesters, Product Marketing Manager, Alan
This highlights how Dust’s influence extends beyond just one team, bridging knowledge silos. Alan estimated over 20% efficiency gains in Sales & Marketing alone.