Armon
Petrossian

Somewhere between the candy he sold out of a backyard treehouse in Portland and the $50 million check from Snowflake Ventures, Armon Petrossian developed a very particular skill: he can look at a Fortune 50 company's data infrastructure and immediately diagnose what's wrong. Not just wrong. Wrong in the specific way he has already solved.

Today, Petrossian is CEO and Co-Founder of Coalesce.io, a cloud-native data transformation platform that has changed how data teams think about ELT workflows. Coalesce gives data engineers a visual, column-aware environment for building Snowflake SQL transformations - with automatic code generation, metadata-driven lineage, and the kind of version control that makes data governance teams stop grimacing at their dashboards.

Emergence Capital, Industry Ventures, and Snowflake Ventures all wrote checks. Inc. named Coalesce a Best Workplace. The company hit 4x ARR growth year-over-year in the fiscal year ending January 2024. None of this happened by accident. It happened because Petrossian spent almost a decade at WhereScape watching data warehousing fail the same companies, in the same ways, for the same fixable reasons.

Petrossian grew up in Portland, Oregon, the son of Armenian immigrants. His father worked as a real estate broker - an immigrant who built something with diligence and consistency. That pattern left a mark. Armon didn't wait for permission to start selling: as a kid he was already running a candy operation out of a treehouse.

He studied at the University of Oregon, double-majoring in Business and Finance while adding Mandarin Chinese - a language he actually went to China to learn, spending a year studying abroad. That willingness to go to the source, to learn the hard thing properly, is a throughline in how he operates.

At 23, he was coaching high school lacrosse, working in tech, and earning decent money. He joined WhereScape as an ETL intern. Over the next nine years, he became National Sales Manager and worked with over 1,000 companies - Fortune 50 clients with sprawling data warehouse problems that kept repeating themselves. He and his future co-founder Satish Jayanthi, who became Coalesce's CTO, were both watching the same failure modes play out at scale.

That year of pure building is part of why Coalesce's product is architecturally different from competitors. They had time to think about column-aware lineage, about how metadata should flow through transformation pipelines, about what a data engineer actually needs versus what existing tools forced them to do. When Coalesce launched publicly in 2022, it didn't feel like a pivot or a minimum viable anything. It felt finished - which in data infrastructure is an unusual sensation.

Coalesce sits inside the cloud data warehouse - natively built for Snowflake, with expanding support for Databricks, Microsoft Fabric, Redshift, and BigQuery. Where other tools abstract the SQL away and make data engineers feel like they've lost control, Coalesce does the opposite: it generates the code, surfaces the lineage, automates the documentation, and lets the engineer see exactly what's happening.

The platform has a visual drag-and-drop interface, but it's not a no-code toy. It's a low-code environment built for engineers who know what they're doing and want to move faster without sacrificing auditability. Column-level lineage. Metadata-driven automation. Automatic SQL generation. Version control that integrates with Git workflows. Governance policies that actually run.

Customers include Caesars Entertainment Group, Houlihan Lokey, CKE Restaurants, Denny's, and TotalEnergies - exactly the kind of organizations where data complexity is a given and transformation errors have real consequences. Petrossian's line says it plainly: "The more complex the data, the better fit they are for Coalesce."

The Series B in April 2024 was a statement round. Snowflake Ventures' participation made explicit what the market already suspected: Coalesce is infrastructure Snowflake wants embedded in its ecosystem. Emergence Capital and Industry Ventures co-led. Bob Muglia (former Snowflake CEO), DNX Ventures, GreatPoint, Next Legacy Partners, Telstra Ventures, and Hyperlink Ventures all came in alongside. That is not a round you close unless the ARR numbers are real.

Petrossian is a sales-trained founder, which means he listens before he pitches. Nine years of WhereScape conversations across 1,000+ enterprises gave him a diagnostic fluency that most CTOs lack. He doesn't describe data transformation as a technology problem. He describes it as an organizational problem that a technology can solve - the bottleneck between raw data and the decisions that raw data is supposed to enable.

He has talked publicly about what it means to build within an ecosystem - specifically Snowflake's - as a strategic choice rather than a constraint. The Think Like An Owner podcast episode where he discussed Coalesce's go-to-market drew on his time in China learning Mandarin as an analogy for immersing yourself in a new context: you don't translate from the outside, you learn to think from within.

He and co-founder Satish Jayanthi co-host The Data T Podcast, a monthly show that brings in industry experts to talk data modeling, DataOps, AI/LLMs, and engineering trends. It's not a vanity podcast. It's a community signal that tells the data engineering world where Coalesce is paying attention.

His leadership philosophy, articulated at the 20/20 Leadership Conference in July 2025: "Real leadership isn't about having all the answers - it's about building systems that don't need you to have them." That's not a fortune cookie. It's the operating principle of a company that has nearly doubled headcount over two years without losing architectural coherence.

Coalesce's Inc. Best Workplace designation in 2024 is the kind of recognition that matters because employees opt into it. You don't get it by surveying executives. You get it by building something that people want to work at - which is a different kind of product problem than the one Petrossian solves on the platform side, but he appears to be running both plays at once.

Coalesce launched as a Snowflake-native platform. The $50M Series B is, in part, the capital to become something bigger: a transformation layer that works across Databricks, Microsoft Fabric, Redshift, BigQuery, and every cloud data platform where enterprise data teams are building. The AI integration is already in motion - AI-powered data lineage, AI documentation assistants, automated metadata enrichment tools are part of the current product roadmap.

The shift from single-platform to multi-cloud is where most data tools stall out. Petrossian's thesis is that the column-aware, metadata-driven architecture Coalesce was built on is platform-agnostic by design - it was never really a Snowflake product, it was a data transformation product that happened to launch in Snowflake's ecosystem first because that's where the pain was loudest.

With Snowflake Ventures in the cap table and Bob Muglia as an investor, Coalesce has both the ecosystem access and the strategic optionality to move fast. Petrossian has been quiet about specific expansion timelines. He tends to build before he announces. That's consistent. It's also why the R&D year mattered: he learned that shipping too early is worse than shipping late, and he's been running that lesson ever since.