Only 15% of organizations are ready to run AI agents in production. The Fivetran-dbt Labs merger is hiring an entirely new role to close that gap.

A Single Data Layer Purpose-Built for AI Agents

On June 1, 2026, Fivetran and dbt Labs closed an all-stock merger announced the previous October, combining two of the most widely adopted data platforms into a single company operating as Fivetran + dbt Labs. The combined entity serves more than 100,000 data teams globally, including teams at OpenAI, Zendesk, HubSpot, Siemens, Roche, and DocuSign, and is led by Fivetran CEO George Fraser with dbt Labs founder Tristan Handy as President.

The rationale is straightforward: the data infrastructure that serves human analysts is not the same infrastructure that serves AI agents. Agents operate continuously, in parallel, and at machine speed. Many organizations are moving toward autonomous agents with no human in the loop, which demands data that is reliable, fresh, governed, and accessible across every enterprise system simultaneously. Fraser put it directly: "The next generation of enterprise AI will be defined by the quality and trustworthiness of the underlying data."

What each company brought to the table was complementary but incomplete on its own. Fivetran moves and transforms data from source systems into warehouses and lakes; it covers the ingestion and synchronization layer. dbt Labs provides the transformation, testing, and semantic modeling layer that defines business logic and enforces data quality. Together, the merged platform spans the full pipeline: from the moment data moves through every transformation to the governed context an agent reasons from. Neither company offered that end-to-end chain alone.

The merger also shipped its first joint product innovations on day one. dbt Core v2.0 (alpha) open-sources the dbt Fusion engine runtime under Apache 2.0. dbt State, a caching layer that skips unchanged pipeline steps, claims infrastructure cost reductions of 30% or more. dbt Wizard (beta) provides autonomous assistance for model authoring and debugging grounded in full project context. And Agents Schema, an open-source standard that designates a single schema in a warehouse as a shared context layer for AI agents, stores metric definitions, semantic models, lineage, and business documentation in plain SQL tables that any SQL-capable agent can read.

The combined company is building on open standards and says it works across any cloud, engine, and tool. Whether organizations actually get vendor-neutral portability or find themselves locked into a new consolidated stack is a question the next few quarters of hiring and product integration will answer.

What New Role Categories Look Like

The combined entity is approaching $600 million in annual recurring revenue with well north of 10,000 customers, according to Business Wire. But the more interesting story is what the company is building that didn't exist before: a unified data layer that moves, transforms, and activates data specifically for AI agents. That architecture is creating job categories that neither company hired for independently.

The core thesis from both CEOs is straightforward. Fivetran's 700 connectors handle ingestion, getting data from SaaS tools, databases, and files into a warehouse or lake. dbt handles transformation, modeling that data into structured, tested, AI-ready outputs. Together, they cover the full pipeline from raw source to agent-consumable context. Fraser called it "open data infrastructure" that works across any compute engine, catalog, BI tool, or AI model. Handy said the merger delivers "the data infrastructure for agents you trust, from that same end-to-end flow to the context an agent reasons from."

That end-to-end scope is what opens up new roles. When ingestion and transformation were separate products at separate companies, each had its own engineering teams, its own solutions architects, its own customer success playbook. Now the combined platform needs people who understand the full data lifecycle as a single system and who can build and support pipelines optimized for how AI agents actually consume data.

The role categories taking shape fall into a few clusters. First, AI data reliability engineering, roles focused on ensuring the data feeding agents is accurate, fresh, and well-governed. Fivetran signaled this direction in May 2026 when it announced plans to become steward of the Great Expectations open source community and GX Core, one of the most widely used data quality frameworks. Owning that project means hiring engineers who can integrate quality checks directly into automated pipelines, not bolt them on after the fact. The 2026 Agentic AI Readiness Index Fivetran released the same month found that only 15% of organizations are fully prepared to support agentic AI in production, a gap that reliability engineers are meant to close.

Second, agent pipeline architecture, roles that design and optimize the end-to-end flow from source data to agent-ready outputs. This is distinct from traditional data engineering because the downstream consumer isn't a dashboard or a report but an AI agent that needs structured, contextual, low-latency data. These architects need to understand both Fivetran's connector ecosystem and dbt's Fusion engine, which powers the transformation layer and is the standard for AI-ready structured data.

Third, solutions engineering for AI-native analytics, pre-sales and post-sales technical roles that help customers design data stacks specifically for AI workloads. Handy told CRN that channel partners are excited about the combined platform because it means fewer vendors to procure and more time spent on high-value AI projects rather than integration plumbing. That shift creates demand for solutions engineers who can speak to both the ingestion and transformation sides and who understand AI use cases like customer 360, regulatory reporting, and real-time agent decisioning.

The combined company's hiring also reflects a broader technical bet: that AI workloads demand new compute patterns. Handy noted that the platform is designed to work not just with the "Big Five" clouds and data platforms (AWS, GCP, Azure, Databricks, Snowflake) but also with engines like DuckDB, Apache DataFusion, and ClickHouse that are "much more tuned for the type of workloads that AI really cares about." Engineers and architects who understand that heterogeneous compute landscape are now in demand at a company that explicitly refuses to lock customers into any single platform.

None of this means the old roles disappeared. Both companies continue to hire for core engineering, and dbt Core remains open source with a dedicated team maintaining it. But the merger created a platform with a scope that neither company had alone, and the hiring is following that scope into territory that didn't exist in either org chart before October 2025.

Why Data Engineering Careers Are Shifting Underfoot

The merger didn't create a new job category out of thin air. It formalized one that had been taking shape across the industry for the past eighteen months, and the hiring now underway at the combined company is the clearest proof yet that "data infrastructure for AI" has graduated from buzzword to actual role descriptions with actual salaries behind them.

To understand how fast this is moving, look at what data engineering work actually consisted of two years ago. Most teams spent the bulk of their time on batch ETL pipelines, schema maintenance, and dashboard support. A 2025 analysis from Data Engineer Academy puts it bluntly: by 2026, the profession's center of gravity shifts from maintaining pipelines to building AI-ready data systems, automated pipelines, real-time analytics, and ML-integrated workflows. That's not a marginal change. It's a redefinition of what the job is.

The AI agent angle accelerates this further. Kestra's 2025 trends report notes that data teams across the industry are experimenting with agentic systems, tools that plan tasks, break them into steps, and execute autonomously. Building pipelines that serve these agents requires a different architectural mindset than building pipelines that feed a Tableau dashboard. The data has to be fresh, well-governed, and structured in ways that an LLM or agent can reason over it directly. That's the gap Fivetran-dbt Labs is now hiring to fill.

What the roles actually look like

The job postings emerging from the merged company map onto a broader pattern visible across the market. Zero G Talent's board shows Databricks adding 67 roles in the past week alone, including Solutions Architects in Berlin and Munich and Delivery Solutions Architects in the US Midwest. These aren't generic data engineering titles. They're roles that assume fluency in both data pipeline architecture and AI/ML integration.

Across the industry, LinkedIn data shows cloud-specific data engineer roles, AWS, Azure, GCP, OCI, now routinely list ML platform expertise (SageMaker, Azure ML, Vertex AI) alongside traditional warehouse and streaming skills. The ITVersity career guide for 2025 lists senior data engineer responsibilities as including designing scalable pipelines "for analytics, ML, or even LLMs." A year ago, that last part wasn't in the job description.

The salary signal

Compensation data confirms the shift. The premium reflects scarcity: engineers who understand distributed systems, can build and operate real-time streaming pipelines, and know how to integrate ML models into production data workflows are rare. The Kestra report frames this as a version of Jevons' Paradox: as AI tools make basic coding cheaper, demand for experienced engineers who can design the underlying architecture actually goes up.

The skill migration

The practical upshot for working data engineers is a concrete set of skills to build now. Data Engineer Academy's 2026 outlook identifies ML concepts for pipeline integration, cloud architecture, real-time processing, and governance compliance as the core competencies that separate engineers who will advance from those who will plateau. The Stack Overflow 2024 survey found PostgreSQL used by 49% of developers, surpassing MySQL for the first time, driven partly by extensions like pgvector that let teams build retrieval-augmented generation workflows without leaving the database. Engineers who can work across that boundary between data infrastructure and AI application layers are the ones commanding the $200K+ offers.

None of this means traditional data engineering is dying. SQL, Python, ETL design, and distributed systems knowledge remain foundational. But the margin, the work that's growing fastest and paying the most, is in the AI-adjacent layer. The Fivetran-dbt Labs merger is just the most visible signal that the industry has decided this is where the next generation of roles lives.

The Go-to-Market Hiring Surge

The merger isn't just reshaping engineering org charts. It's triggering a parallel hiring wave on the commercial side, one that reveals how seriously the combined company believes the merged platform can sell itself into enterprise AI budgets.

Scanning the open roles on both careers pages, the pattern is hard to miss. Fivetran's board lists multiple Senior Sales Engineer positions across commercial and enterprise segments, with locations in Denver, Dublin, London, and remote Germany. There's also a Lead Sales Engineering Specialist, Security role spanning Ireland and the UK, a title that didn't exist at either company before the merger. dbt Labs, meanwhile, is hiring Customer Solutions Architects in Austin, a Staff Solutions Architect in Dublin, and a Solutions Architect, Enterprise/Majors for the US West region. The dbt Labs board also shows a Manager, Customer Solutions Architect role in Dublin, a leadership position overseeing what is clearly a growing pre-sales function.

These aren't generic sales hires. The titles and role descriptions point to a go-to-market motion built around a specific technical story: ingestion plus transformation, sold as a single AI-ready data layer. A solutions engineer pitching that combined platform needs to understand both Fivetran's connector ecosystem and dbt Labs' transformation and semantic modeling capabilities. That's a different skill set than what either company's pre-sales team carried independently.

The business development side shows the same specialization. Fivetran is hiring Business Development Representatives segmented by enterprise and commercial tiers, including a French-speaking enterprise BDR based in London and a commercial BDR in Japan. dbt Labs has an Enterprise Sales Director for Australia and a Japan Sales Director, both remote, roles that suggest the combined company is pushing into geographies where neither brand had a standalone enterprise sales presence.

Against the backdrop of the enterprise AI market's rapid growth, the combined entity is staffing up to capture budget that's shifting from traditional BI and analytics tooling toward AI data infrastructure. IDC analyst Devin Pratt noted that 97% of organizations want to reduce the number of products they use for data management, a dynamic that favors a merged platform over point solutions, but only if the sales team can articulate the integration story convincingly.

The competitive pressure is real. Zero G Talent's board data shows Databricks adding 67 roles in the past week alone, including Solutions Architect positions in Berlin and Munich and a Senior Director, Global Accenture Lead in the US. That's a company with a similar enterprise AI data platform story, hiring aggressively in the same pre-sales and solutions engineering roles. The talent pool for people who can sell and technically position AI-native data infrastructure is finite, and both companies are fishing from it.

For candidates with solutions engineering or BDR experience in data platforms, the merged Fivetran-dbt Labs entity is now one of the more visible employers, and the roles are more specialized than what the market offered even six months ago.

Snowflake, Databricks, and the Talent War for AI Data Skills

The merger didn't happen in a vacuum. Both companies already sat inside a data stack that Snowflake and Databricks are aggressively fighting to own, and the combined entity is now competing directly with those platforms for the same engineers, solutions architects, and go-to-market specialists.

Start with the overlap. A LinkedIn analysis of 2025 data engineering job postings found that Snowflake, dbt, and Fivetran together form what many teams treat as the default modern data stack, with engineers fluent in all three commanding an 18% compensation premium over peers who aren't. That premium is the market telling you something: the skills the merged company needs are the same ones Snowflake and Databricks are hiring for at scale.

The platform-level ties run deep. dbt Labs was named Snowflake's 2026 Data Integration Product Partner of the Year, the fourth consecutive year Snowflake has given dbt Labs a partner award. Over 75% of Snowflake account customers use dbt, and 90% of joint customers actively run Snowflake Cortex AI. The two companies co-hosted a June 2025 webinar on building AI-powered data pipelines together, with Snowflake's own principal solution engineer presenting alongside dbt Labs staff. Snowflake's interest in this talent pool isn't passive: the company acquired Select Star, a data catalog and lineage tool that integrates with both dbt and Fivetran, adding another surface where these skills converge.

Databricks is playing the same game from the other side. Fivetran showed up at the Databricks Data & AI Summit in June 2025 with a booth and four speaking sessions, including a hands-on lab walking attendees through building pipelines with Fivetran, dbt Cloud, and Sigma on top of Databricks. The session featured staff from both Fivetran and dbt Labs, a preview of the combined company's go-to-market motion before the merger was even announced. Today, Zero G Talent's board lists 67 roles added at Databricks in the past seven days alone, spanning solutions architecture, product program management, and compensation leadership.

The talent war has a price floor. Engineers who can design pipelines that feed reliable, well-modeled data into agent systems are rarer than generalist ML engineers, and the merger is effectively creating a new job title around that scarcity.

For candidates, the leverage is real. The merged company needs people who understand Fivetran's connector ecosystem and dbt's transformation layer as a single system, a skill set that Snowflake, Databricks, and a handful of startups are also chasing. The engineers and solutions architects who fit that profile aren't choosing between one employer. They're choosing between an entire ecosystem that's competing for their attention.

Where the Combined Company Is Hiring

The merged entity is pulling talent from a wider map than either company covered alone. Fivetran lists nine office locations on its careers page, Oakland (headquarters), Denver, Bengaluru, Dublin, München, Amsterdam, London, Sydney, and Novi Sad, while Built In confirms five physical offices: Oakland, Denver, Bengaluru, Dublin, and Novi Sad. dbt Labs runs a distributed-first model, with open roles spanning Austin, Dublin, and remote positions across the US, India, Ireland, France, Germany, the UK, Australia, Japan, and Canada.

The overlap tells you where the combined company is concentrating its hiring muscle.

Oakland and Denver are the twin US hubs. Fivetran's careers page shows the densest cluster of engineering, product, marketing, and systems-engineering roles split between these two cities. A Sr. Principal AI Systems Architect sits in Oakland. Staff Site Reliability Engineers are listed in both Oakland and Denver. Product managers for observability, reverse ETL, and enterprise platform roles are duplicated across both locations. Denver also anchors a large share of the go-to-market hiring, business development representatives, account directors, and sales engineers for commercial and enterprise segments.

Bengaluru is the engineering backbone. Fivetran lists more than a dozen open engineering roles there, from Staff Software Engineers to Senior SDETs. dbt Labs adds senior infrastructure and platform software engineers in India (remote, likely Bengaluru-based). If you're a mid- or senior-level data platform engineer, this is the single largest concentration of open requisitions in the combined company.

Novi Sad is the quiet workhorse. Fivetran has built a substantial engineering outpost there, Senior C Programmers, Database Connector engineers, Staff DevOps Engineers, and a Staff Site Reliability Engineer all list Novi Sad as the location. It's a lower-cost European hub handling core database and infrastructure work, and the headcount suggests it's growing, not static.

Dublin and London anchor the European go-to-market push. Fivetran lists product marketing managers in both cities. dbt Labs has a Staff Solutions Architect and Manager of Customer Solutions Architects in Dublin, plus enterprise account executives and sales directors in London and Dublin. The combined entity is clearly using these two cities to cover EMEA sales and solutions engineering.

Austin is dbt Labs' US commercial hub. Four open roles, two account executives, a customer sales director, and a solutions architect, are based there, with no corresponding Fivetran listings. This is where dbt Labs' sales culture is concentrated, and post-merger, it becomes the combined company's second US commercial node behind Denver.

The rest is remote and distributed. dbt Labs lists roles open to remote workers in the US, Canada, France, Germany, Ireland, the UK, Australia, and Japan. Fivetran has remote roles in India, Japan, and scattered US states. This isn't a remote-first company, but it's not pretending geography doesn't matter either; the hybrid model, with two days a week in-office at Fivetran's locations, is the stated norm.

The broader AI infrastructure job market backs up these choices. Nationwide, employment tied to AI and data center infrastructure reached roughly 483,000 jobs in 2025, according to the Bureau of Labor Statistics. Oakland and Denver both rank among the top US cities for data center and AI infrastructure roles. Bengaluru remains the default offshore engineering hub for US data companies. Novi Sad has emerged as a European engineering talent pool that most US hiring managers overlook, and Fivetran is not the only data infrastructure company building there, but it's one of the most visible.

For candidates, the implication is straightforward: if you're in Oakland, Denver, Bengaluru, or Novi Sad, the combined company is hiring at volume. If you're in Dublin or London, the roles skew toward solutions architecture and enterprise sales. Everywhere else, remote is an option, but the center of gravity is still physical.

Working in AI? Zero G Talent tracks the openings: browse AI jobs, openings at Databricks, and the people building the field.