Sifflet 2025: The Year Data Observability Became a Business Imperative

2025 was a defining year for Sifflet. We shipped AI agents that fundamentally changed how teams monitor data, expanded into new industries, and closed a new funding round to fuel our North American expansion. Here's how it happened.

New Funding, New Industries, New Scale

Our existing investors  EQT Ventures and Mangrove Capital Partners doubled down on Sifflet in a new funding round, with Capmont Technology joining as a new investor. The capital is powering our push into North America, where enterprise demand for AI-ready data infrastructure has accelerated faster than anyone predicted.

More importantly, we expanded into industries where data reliability isn't optional, it's a regulatory requirement. Leading financial services, insurance, and logistics & shipping companies now run on Sifflet. These are sectors where a data incident doesn't just cause inconvenience; it triggers compliance risk and operational disruption. The fact that they chose us validates our approach: treating data quality as both a technical and business problem.

Usage That Proves Product-Market Fit

The numbers that matter most to us aren't funding figures, they're adoption metrics.

This year we crossed 5,000 users on the platform. We saw 7x growth in monitors created and 10x growth in monitor runs. That's not just more customers; it's existing customers going deeper. Two things drove this:

First, we introduced automatic incident grouping, a game changer for incident management. Instead of drowning teams in individual alerts, Sifflet now uses AI and data lineage to automatically group related monitor failures into a single incident. The system analyzes relationships between monitors freshness, volume, schema changes and validates connections using lineage. The result: less alert fatigue, faster root cause identification, and teams that actually trust their monitoring.

Second, Sentinel changed who can create monitors and how. Our AI agent analyzes metadata and recommends comprehensive monitoring strategies no SQL required, no guesswork about which tables need coverage. Data analysts and business users who previously relied on engineers are now spinning up monitors themselves. Democratizing monitor creation unlocked usage we couldn't have achieved with a technical-only approach.

Doubling Down on Business-Context Aware Observability

This year we sharpened our positioning around what actually differentiates Sifflet: business-context aware observability.

Most data observability tools focus purely on technical metrics row counts, schema changes, freshness and technical users i.e. the data engineers. That's necessary but not sufficient. A 5% drop in row count might be catastrophic for one table and completely normal for another. Context is everything.

Sifflet connects technical data health to business impact. We help teams understand not just what broke, but why it matters and who needs to know. That's the difference between alert fatigue and actionable intelligence.

New Leadership to Match Our Ambition

We made key hires to scale our go-to-market and operations:

Joe Steadman joined as Head of Sales to lead revenue growth. After nearly 10 years at Matillion from 1 to 100mln ARR, he is well equipped to build the commercial engine for our next phase.

Rémi Bastien (ex-Contentsquare) joined to lead global operations.

We also opened our New York office, establishing a US headquarters to serve the North American market directly. Our fastest-growing region now has boots on the ground.

And critically, we revamped post-sales and customer success to sustain our new level of scale. More customers means nothing if we can't make them wildly successful.

Product: The Year of AI Agents

Our biggest bet of 2025 was shipping three AI agents that transform how data teams approach reliability:

Sentinel analyzes your metadata and recommends precise monitoring strategies. It went GA in October, and teams are now generating comprehensive monitors in minutes instead of days. The impact on adoption has been massive.

Sage recalls past incidents, understands lineage, and surfaces root causes in seconds. Think of it as institutional memory for your data stack.

Forge suggests contextual fixes grounded in historical patterns. It's not just telling you what broke it's proposing how to fix it.

Former Gartner VP Analyst Sanjeev Mohan called it "a meaningful evolution in data observability... helping teams move from alert fatigue to intelligent, context-aware resolution."

Platform Updates That Actually Matter

Beyond AI, we shipped capabilities our enterprise customers were asking for:

Databricks Workflows Integration: End-to-end visibility into your data pipelines. See which jobs create or update your tables, directly in your lineage.

Revamped Lineage Experience: Completely rebuilt for clarity. New "transformation" nodes show pipeline steps like Airflow DAGs. Better navigation for complex graphs. Toggle between classic and new views.

Subdomains: Hierarchical organization and granular access control. Marketing can now have Brand Marketing and Digital Marketing as subdomains, each with specific permissions.

Conditional Monitors Upgrade: Full feature parity with all other monitors. Advanced joins, incremental scans, group_by, where clauses, and threshold settings.

Domain Management APIs: Full lifecycle management of domains via API. Terraform provider coming soon.

Enterprise Customers Across New Verticals

This year we welcomed customers across financial services, insurance, logistics, and shipping industries where data reliability is table stakes for compliance and operations.

We also deepened relationships with existing customers like Penguin Random House and Saint Gobain. Pete Williams, their Chief Data Officer, captured what makes Sifflet different: "Their AI-native approach doesn't just detect issues, it gives our team the context to act faster and smarter."

The common thread? Companies no longer see data observability as a nice-to-have. It's the foundation that makes GenAI possible.

G2 Recognition

Users ranked Sifflet Best Estimated ROI and Most User friendly in the data observability category. Twenty-seven new badges this year. We'll take the customer proof over marketing claims any day.

Signals Summit: Our First Major Conference

In November, we hosted Signals Summit 2025: Trust by Design, our first major industry conference focused on data reliability in the AI era.

The lineup spoke for itself: Tristan Handy (dbt Labs), Joe Reis, Michel Tricot (Airbyte), Sanjeev Mohan, and data leaders from Penguin Random House, Snowflake, and dozens of enterprises shared hard-won lessons on building trustworthy data foundations.

The theme designing for reliability, not bolting it on resonated. Thousands of data practitioners joined across four days of sessions on metadata, governance, lineage, and what it actually takes to make AI dependable in production.

We're already planning the next one.

What 2026 Looks Like

We're doubling down on three bets:

1. AI agents going deeper Sage and Forge capabilities will expand. The goal is proactive resolution, not just detection.
   
   
2. North American expansion The US is now our fastest-growing market. Expect a larger presence on the ground.
   
   
3. And one more thing we're not quite ready to announce.…but let's just say we're rethinking how teams discover and adopt Sifflet. Stay tuned!
   
   

The market has caught up to what we've believed since day one: data quality is both an engineering and a business problem. The 7x and 10x growth in monitors and runs tells us we're solving it the right way. 2026 is about scaling what works.

Thanks to our customers, investors, and team for an exceptional year.

Salma Bakouk, Co-Founder & CEO