How Brevo automated go-to-market workflows with Dust & Supabase: 80% faster personalization, zero engineering tickets
- Industries
- TechnologyB2B SaaS
- Company Size
- 201-1000
- Departments
- SalesMarketingOperations
Alexandre Le Goupil
Head of Revenue Systems and AI
Key Highlights
- 80% time reduction on email personalization (30+ minutes → minutes per prospect)
- 30%+ of support requests automated through self-serve AI agents
- 2,500+ production actions executed through Supabase-connected agents since summer 2025
- Days, not months from concept to production workflows
"Dust turned our RevOps team into builders. We went from zero AI in production to thousands of automated actions — and we never had to write a ticket to engineering.”— Alexandre Le Goupil, Head of Revenue Systems and AI, Brevo
About Brevo
Founded in 2012, Brevo is a Paris-based Omnichannel Customer Engagement platform serving over 600,000 customers globally. With over 1,000 employees, Brevo achieved unicorn status in December 2025. The company surpassed $218M in ARR in 2025 while maintaining double-digit EBITDA margins.
What sets Brevo apart? They practice what they preach: using AI extensively within their own sales and marketing teams to deliver hyper-personalized outreach at scale.
The Challenge: Personalization at scale hit a wall
Brevo's Revenue Operations team faced a problem familiar to any scaling go-to-market organization: their processes couldn't keep up with their ambitions.
The personalization bottleneck:
Sales reps needed to send hyper-personalized emails to hundreds of prospects weekly. Generic outreach no longer worked. But manually researching each prospect, combing through LinkedIn, the company website and CRM history took 30+ minutes per person. At that pace, personalization meant sacrificing volume.
Data silos everywhere:
Critical intelligence lived scattered across systems. CRM data sat locked in their CRM system. Product documentation lived in Notion. Communication history was buried in Slack. Web research required manual searches. Sales reps spent more time hunting for context than actually selling.
RevOps as a bottleneck:
An estimated 30%+ of internal requests to Revenue Operations could be answered with existing documentation: "Which customers are in e-commerce?", "What's our pricing for this segment?", "Who do we know at this company?", but required manual human responses every time. This pulled the RevOps team away from strategic work.
No scalable referral process:
Finding relevant customer examples to share with prospects required manual CRM searches and tribal knowledge. One sales rep might know which customers to reference for e-commerce prospects, another for SaaS. There was no systematic way to surface this intelligence.
"Salespeople would come to me asking for a list of our best e-commerce customers before a call. Or they'd need to know if we had history with a prospect's company. That context was all in the CRM, but our AI agents were completely blind to it. I needed to connect those two worlds."— Alexandre Le Goupil
The cost wasn't just time. It was missed revenue moments and inconsistent execution across the sales team.
The Turning Point: What if RevOps could ship like engineers?
Alexandre Le Goupil and the Revenue Operations team had already adopted Dust to build AI agents. The agents were good at research, reasoning, and writing. But the CRM was completely out of reach, and that data was what would make the agents' output highly actionable.
The team needed more than read-only access to dashboards. They needed ****agents that could both read contextual data from their systems AND write AI-generated outputs back into production workflows.
Most importantly, they needed to move fast. Without waiting for engineering or complex processes.
Other Dust customers were using BigQuery, Snowflake, and Databricks for their data platforms, but none of those were built for what Alexandre needed: a database that AI agents could query and update in real-time, through a standard protocol.
The Solution: Dust + Supabase as the operational data foundation
The connection came through a conversation between the Dust team and Pedro Rodrigues, AI Engineer at Supabase. Supabase is an open-source backend platform built around managed PostgreSQL databases — and one of the earliest database providers to adopt the Model Context Protocol (MCP). Pedro presented the Dust team with a production-ready MCP server that could be integrated in just a few clicks.
What made it the right fit for Brevo? The MCP is fully hosted by Supabase with complete read and write capabilities. No infrastructure to manage, no servers to maintain — just a live operational database that Dust agents could query and update in real-time, through a standard protocol, without custom integrations for every workflow.
For Dust users, connecting to Supabase was already simple. But to make things even easier, Dust added Supabase to its official MCP integrations list in October 2025, making it accessible with a single click.
"BigQuery and Snowflake are great for analytics. But I needed something an AI agent could query live, in the middle of a conversation. Supabase gave me Postgres with an MCP server ready to go. That was what decided it."— Alexandre Le Goupil
Implementation took days, not months:
Brevo started by connecting to Supabase's remote MCP server before Dust launched its official one-click integration. When the official version shipped in October 2025, they migrated immediately. The technical setup took minutes. The real work was documenting their data schemas so Dust's agents would understand what each field meant and which tables to query for which questions.
"Connecting Supabase was the easy part. The real work was writing good documentation for the AI — telling it what each field means, which tables to query for which questions. Once we got that right, everything clicked."— Alexandre Le Goupil
What makes this different:
This isn't a data warehouse for analytics. It's an operational database that Dust agents interact with in real-time to power production workflows. Supabase serves as the synchronization hub between Brevo's CRM, Dust agents, and their go-to-market systems.
Supabase was the first data platform where Dust enabled both read and write capabilities through natural language for AI agents — representing what's possible when database providers and AI platforms work together through open standards like MCP.
Three Production Workflows That Transformed Go-to-Market
Workflow 1: Customer Referral Finder — Instant social proof for every sales call
The business need:
Sales reps needed relevant customer examples before prospect calls to demonstrate social proof and credibility. Manually searching the CRM took 10-15 minutes and often missed the best matches.
What the agent does:
Sales reps now ask a Dust agent: "Find me the top 3 e-commerce customers we can reference for this prospect." The agent queries the CRM mirror in Supabase by industry, company size, and product usage, then retrieves contextual details—implementation timeline, key results, specific use cases.
The bidirectional loop:
- Reads: CRM data (Companies, Contacts, Deals) stored in Supabase
- Outputs: Structured recommendations with customer context ready to use in conversations
Impact:
What previously took 10-15 minutes of manual CRM searching now happens in seconds. Every sales rep gets the same quality of intelligence, eliminating the inconsistency that came from tribal knowledge.
"That workflow alone changed how reps prepare for calls. They get reference customers, account context, and a suggested angle in seconds. It freed up hours of my team's time every week."— Alexandre Le Goupil
Workflow 2: Automated Personalized Email Generation — Hyper-personalization at scale
The business need:
Business Development Representatives needed to send personalized email sequences to hundreds of prospects weekly, but manual personalization was unsustainable at that volume.
What the agent does:
BDRs select contacts to prospect in their CRM. This triggers a Dust conversation that:
- Pulls the prospect's full history from Supabase (prior contact, pipeline stage, products discussed)
- Enriches with web context (LinkedIn profiles, company news, firmographic data)
- Routes to specialized sub-agents based on prospect type (gated content downloads get one agent, cold e-commerce leads get another)
- Generates 3 personalized emails tailored to the person's role, seniority, and context
The bidirectional loop:
- Reads: CRM history from Supabase + web/LinkedIn data via Dust
- Writes: Generated email sequences (as structured JSON and HTML) back to Supabase
- Production use: Brevo's CRM pulls these emails into multi-channel sales sequences (email, phone, LinkedIn)
Anti-hallucination safeguards:
The team built strict rules into their prompts: only reference data that exists in the database. Grounding every agent in structured Supabase data eliminated the risk of fabricated customer names or made-up statistics that would destroy trust.
Impact:
- 80-90% time reduction: From 30+ minutes of research and writing per prospect to minutes
- Quality improvement: AI-generated emails are consistently more detailed because agents access complete CRM history, web research, and customer examples simultaneously—context no human could feasibly gather for every prospect
- Consistent execution: All sales reps now have access to the same personalization capabilities, eliminating tribal knowledge gaps
"We used to send the same generic email to every e-commerce prospect. Now every email reflects who this person is and what we know about them. Supabase holds the context going in and stores the output coming out. It's the connective tissue between our CRM and our AI."— Alexandre Le Goupil
Workflow 3: Personalized Landing Pages on Demand — Dynamic marketing at scale
The business need:
Brevo's marketing team needed a way to deliver personalized guidance to prospects at scale, without manual work for each visitor.
What the agent does:
When a visitor submits their email and company name on a Brevo landing page, the data is stored in Supabase and triggers a Dust agent. The agent generates a tailored marketing plan for that company — channel mix recommendations, campaign ideas, and timelines — then writes the output back to Supabase, which renders it as a unique, personalized page for that visitor.
The bidirectional loop:
- Reads: Visitor-submitted data (email, company name) stored in Supabase
- Writes: Generated marketing plan back to Supabase
- Production use: Supabase renders the output as a dynamic, personalized page for each visitor — in seconds, with no manual intervention
Impact:
- Instant personalization at scale: Every visitor gets a custom marketing plan the moment they submit a form
- Same architecture, new team: The Dust + Supabase stack built for sales now powers marketing lead generation
- Fully automated end-to-end: From form submission to personalized page render, no human in the loop
"Someone fills out a form and gets a custom marketing plan in seconds. Data goes into Supabase, the AI builds the plan, the result comes back through Supabase as a page. We kept the whole thing as simple as we could."— Alexandre Le Goupil
Results: From manual operations to intelligent automation
Since implementing Dust + Supabase, Brevo's Revenue Operations team has fundamentally transformed how their go-to-market teams work:
Quantitative impact:
- 80% time reduction on email personalization (30+ minutes → minutes per prospect)
- 30%+ of internal support requests now handled by self-serve agents
- 2,500+ production actions executed through the Supabase MCP since June 2025
- Days, not months from concept to production workflows
- Batch processing capability: Email generation for multiple prospects simultaneously, enabling campaign launches that previously took days
Qualitative transformation:
- Higher quality personalization: AI-generated content is more comprehensive because agents access complete context simultaneously
- Reduced cognitive load: Sales reps simply ask agents in natural language rather than remembering where data lives
- Faster iteration: RevOps tests new workflows rapidly without engineering dependencies
- Confident in outputs: Agents pull from verified Supabase sources rather than hallucinating, ensuring accuracy
- Consistent sales experience: All reps access the same intelligence, eliminating tribal knowledge gaps
- Operational reliability: Since connecting Supabase, the team hasn't had to troubleshoot, reconfigure, or debug the connection. It just runs.
"We set it up, it worked, and we haven't gone back. It just runs. That's exactly what you want from infrastructure when you're a small ops team trying to ship fast."— Alexandre Le Goupil
What This Enabled: An ops team that ships like engineers
Brevo's Revenue Operations team isn't an engineering team. They don't write backend services or manage infrastructure. But with Dust + Supabase, they built three production AI workflows that run daily across Sales and Marketing, without filing a single ticket with engineering.
This is the pattern emerging across forward-thinking revenue organizations: when you give non-engineering builders production-grade tools with AI-native capabilities, they stop waiting and start shipping.
The new operating model:
- Question: What new workflow do we need?
- Answer: What agent do we build in Dust?
- Result: Ship in days, iterate based on feedback, scale when validated
What's Next: Expanding the automation playbook
Brevo's team views their current implementation as just the beginning:
Batch processing optimization: Scaling email generation workflows to handle thousands of prospects in single batch runs with improved latency and reliability.
Cross-department expansion: While Sales and Marketing are power users, other departments are beginning to explore use cases based on the success they've seen.
"We started with one use case. Now we have three in production and more on the way. Every time we have a new idea, the first question is: what table do we need in Supabase? That's how fast we can move now."— Alexandre Le Goupil
Why This Matters: The future of Revenue Operations
Brevo's story illustrates a fundamental shift in how modern revenue teams operate. Revenue Operations is no longer just reporting and tool administration. It's building production workflows that directly impact revenue.
The combination of Dust (for AI orchestration and agent creation) and Supabase (for operational data infrastructure) creates a new capability: business teams can ship intelligent automation without waiting for engineering cycles.
Key lessons for other RevOps teams:
- Start with real pain points — Build agents for specific, repetitive workflows that consume significant time, not generic AI experiments
- Document your data thoroughly — Well-defined schemas and field descriptions dramatically improve agent accuracy
- Leverage bidirectional capabilities — Real power comes from agents that write outputs back into production systems, not just read data
- Plan for automation from day one — While manual chat agents are useful, ROI comes from programmatic usage via API
- Expect iteration — Build agents, deploy quickly, gather feedback, refine prompts. Don't aim for perfection before launch.
Interested in learning more about how Dust can help your team? Visit our solutions page or reach out to our sales team.
