Soda
Soda (Brussels, founded 2018) provides AI-native data quality and observability software for enterprise data teams, with open-source Soda Core, contracts, agentic Soda Cleanse for remediation, and 60+ warehouse integrations.
- Company typePrivate
- Founded2018
- HeadquartersBrussels, Belgium
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What Soda does
Soda is a Brussels-headquartered data quality and observability software company founded in 2018, serving enterprise data teams that need to monitor, validate, and remediate data pipelines at scale. The platform centers on Soda Core (open-source) for checks, Soda Cloud for collaboration and observability, and data contracts that codify producer-consumer data agreements. Architecture integrates with 60+ data sources including warehouses, lakes, and catalogs via check execution agents, exposing results through a unified observability layer. AI capabilities are differentiated: the company has published peer-reviewed research at NeurIPS, JAIR, and ACML in collaboration with INRIA, including TabICLv2 for tabular anomaly detection and proprietary validation metrics reported to reduce false positives by 70% versus Facebook Prophet. Newer AI-native products include Soda Cleanse (agentic AI for autonomous data remediation), Contract CoPilot/Autopilot for contract authoring, and an MCP endpoint positioning Soda within the agent tooling ecosystem.
The business model combines a free open-source tier, a Team plan at $750/month, and custom-priced Enterprise annual contracts, with a usage-based SPU (Scan Processing Unit) metering layer introduced for Soda Cleanse. Customer logos span gaming (2K Games), food delivery (HelloFresh), travel (CarTrawler), and insurance (Group 1001), indicating vertical diversification. Soda has raised approximately $31.5M across four disclosed rounds between 2019 and 2024, with lead investors including Point Nine Capital and Singular. The June 2025 acquisition of NannyML extends the company into ML model monitoring, and the company operates with 11-50 employees across Belgium and additional geographies.
Soda firmographics
Firmographics- Name
- Soda
- Legal name
- Soda Data Inc.
- Website
- https://soda.io
- Company type
- Private
- Founded year
- 2018
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Soda (Brussels, founded 2018) provides AI-native data quality and observability software for enterprise data teams, with open-source Soda Core, contracts, agentic Soda Cleanse for remediation, and 60+ warehouse integrations.
- Ownership category
- akta.pro rank
Where Soda is headquartered
LocationHeadquarters
- HQ city
- Brussels
- HQ country
- Belgium
- HQ region
- Europe
Offices1 record
Markets served
Soda business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Marketing or Sales, Infrastructure, Operations
Revenue model
- Freemium tier: Free tier with limited free Soda Processing Units (SPUs) including pipeline testing, metrics observability, alerting, and ticketing integrations; no credit card required, used as a self-serve funnel into paid plans.
- Team subscription: Team plan priced at $750 (per month context), includes collaborative data contracts, no-code interface, advanced AI-powered data quality features, audit logs, custom roles, RBAC, private deployment, SSO, and premium support.
- Enterprise subscription: Custom-priced Enterprise tier for business collaboration; annual billing and volume discounts available; bundles enterprise-grade governance, AI features, and premium support.
- Soda Processing Units (SPU) overage: Usage-based pricing on additional SPUs beyond the free allowance on Team and Enterprise tiers; consumed when running data quality checks at scale.
- Soda Cleanse add-on: Soda Cleanse is an add-on to Soda Cloud (Enterprise plan) available in Private Preview; agentic data cleansing for entity normalization, imputation, deduplication, and reconciliation.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Freemium | Monthly | Free — $0/month with free SPUs |
| Subscription | Monthly | Team — $750 with collaborative data contracts |
| Subscription | Annual | Enterprise — Custom pricing with full governance |
| Subscription | Annual | Soda Cleanse add-on (Enterprise) |
Go-to-market motion5 records
Distribution channels5 records
Marketing channels10 records
Soda product offering
Product offeringCore offering
Soda provides an AI-native data quality platform that detects, explains, and resolves data issues across enterprise data pipelines. The platform combines Soda Core (open-source YAML/SodaCL testing engine) with Soda Cloud (hosted SaaS control plane) and Soda AI (proprietary AI capabilities), delivering data observability, collaborative data contracts, record-level anomaly detection, smart alerting, a Diagnostics Warehouse, and agentic data cleansing. The offering is sold as free, Team ($750), and custom Enterprise subscriptions, with usage-based Soda Processing Units (SPUs) and an add-on for Soda Cleanse agentic remediation.
Product overview
Soda offers a unified, AI-native data quality platform delivered as Soda Cloud (hosted SaaS at cloud.soda.io), with Soda Core as the open-source code-based counterpart for engineers. The platform is organized as a set of interlocking modules that sit on a single shared contract and diagnostics foundation: Data Observability provides the metrics and record-level anomaly detection layer; Collaborative Data Contracts define executable expectations in code or UI; Smart Alerting routes accurate, owner-targeted incidents; Diagnostics Warehouse stores every failed record inside the customer's own environment; Lineage and Impact provides bi-directional catalog/governance integration; Pipeline Fixing shifts checks into CI/CD and orchestration; Soda AI layers Contract Autopilot, Contract Copilot, and the Soda MCP/CLI/API on top; and Soda Cleanse (Agentic Data Cleansing) extends the same contract-driven workflow into remediation through specialized AI agents. Together these modules cover detect–govern–alert–fix as one workflow, with Soda 4.0 as the current unified release.
Differentiator
Problem solved
Functional benefit
Products and services
- Soda AI AI layer for the Soda data quality platform that drafts and refines data contracts (Contract Autopilot and Contract Copilot), exposes an MCP/CLI/API for engineers and agents, and runs on metadata rather than raw rows. Underpinned by peer-reviewed research (NeurIPS, JAIR, ACML).
- Soda Cloud Hosted SaaS platform (cloud.soda.io) bundling Data Observability, Data Contracts, Smart Alerting, Diagnostics Warehouse, Lineage, and the Soda Agent for engineers; base on which Soda AI and Soda Cleanse are layered.
- Soda Data Observability
- Collaborative Data Contracts Executable, version-controlled data contracts letting engineers work in Git and business users in the UI; Contracts Copilot turns plain-English rules into contract language and Contracts Autopilot derives contracts from production data.
- Smart Alerting AI-powered alerting that automatically routes accurate, contextual incident notifications to the right owner through Slack, Teams, Jira, ServiceNow, PagerDuty, or Opsgenie, reducing false positives.
- Diagnostics Warehouse (Record-Level Diagnostics) Captures and stores every failed record directly in the customer's own warehouse, providing queryable history for faster debugging and root-cause analysis without data leaving the environment.
- Lineage and Impact Bi-directional catalog and lineage integration that turns governance rules into executable data quality checks and feeds quality results back to catalog tools (Collibra, Atlan, Purview, Alation, etc.) for unified lineage visibility.
- Soda Cleanse (Agentic Data Cleansing) Private-Preview add-on to Soda Cloud using specialized AI agents (entity normalization, imputation, deduplication, reconciliation) to detect bad records, generate fixes, and write corrections back to source after steward approval. Requires Soda 4.0 and Diagnostics Warehouse (Enterprise plan).
- Pipeline Fixing (Automated Pipeline Fixing) Shift-left data quality integrating Soda checks directly into development and CI/CD pipelines to detect schema changes, reconciliation mismatches, missing values, and anomalies before they reach production; routes fixes into operational workflows.
- Soda Core Open-source, YAML/SodaCL-based data quality framework that engineers run as code in orchestration workflows (Airflow, Dagster, dbt); foundation for Soda Cloud deployments, used in Group 1001 and 2K Games stacks before upgrading to Soda Cloud.
Quantifiable outcome
- Metrics monitoring algorithms deliver 70% fewer false positives than Facebook Prophet and scale to 1B rows in 64 seconds
- +4 more outcomes
Companies that use Soda
Customer profileNamed customers4 records
Segments7 records
Ideal customer profiles5 records
Soda technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration53 records
AI capability11 records
Feature8 records
Soda partnerships and signals
Strategic signalPartnerships
Ten partnerships are on record, tiered major.
- NannyMLmajorIn June 2025, Brussels-based Soda acquired Leuven-based NannyML, an AI monitoring and observability company, to enhance its data quality platform by integrating NannyML's AI technology for detecting data issues, model performance, and data drift — supporting Soda's expansion into AI and ML infrastructure.
- CollibramajorCollibra is listed as a data catalog integration on soda.io/integrations. Soda offers bi-directional sync with Collibra so governance teams can define policies in Collibra while Soda executes them as automated data quality checks and feeds quality results back for unified lineage visibility.
- AtlanmajorAtlan is listed as a data catalog integration. Soda bi-directionally syncs metadata and quality results with Atlan so users can see Soda's quality metrics within the Atlan catalog and link observability to data lineage, as demonstrated in Group 1001's deployment.
- SlackmajorSlack is a primary alerting and collaboration integration; Smart Alerting automatically creates, assigns, and tracks incidents from failed checks in Slack so the right person is notified where they already work.
- AirflowmajorApache Airflow is supported as an orchestration integration; Soda checks can be triggered from Airflow DAGs (as deployed by HelloFresh and CarTrawler) to validate pipelines and stop refreshes when data quality fails.
- SnowflakemajorSnowflake is one of Soda's primary data sources; Group 1001 and CarTrawler run Soda checks directly against Snowflake for validation, reconciliation, and feature-store quality.
- DatabricksmajorDatabricks is supported as a core data source; 2K Games runs Soda observability across Databricks-hosted game telemetry pipelines alongside Snowflake.
- Amazon Web Services (AWS)majorAWS is listed as an infrastructure integration; Soda runs natively on AWS and supports running data quality checks against AWS-hosted data sources and warehouses.
- Microsoft AzuremajorMicrosoft Azure is listed as an infrastructure integration; Soda supports Azure-hosted deployments and integration with Azure Active Directory for SSO.
- Google Cloud Platform (GCP)majorGoogle Cloud Platform is listed as an infrastructure integration; Soda supports GCP-hosted deployments and integration with BigQuery.
Scale indicators8 records
Recent moves6 records
Expansion highlights6 records
Soda competitors and assessment
Company assessmentDirect peers
- Monte Carlo Data: Monte Carlo is a direct peer offering a data observability and reliability platform with anomaly detection, freshness, and schema monitoring; it competes head-to-head with Soda Data Observability for enterprise data team budgets.
- Bigeye: Bigeye is a direct peer providing automated data quality monitoring, anomaly detection, and metrics across data warehouses; it targets the same data engineer and analytics buyer as Soda with a similar SaaS plus enterprise motion.
- Great Expectations: Great Expectations is a direct peer in the open-source data quality testing category alongside Soda Core, offering code-based expectation suites for pipelines; it competes for developer mindshare and feeds commercial offerings that overlap with Soda Cloud.
- dbt Labs: dbt Labs is a direct peer through dbt tests (its built-in data testing framework) and dbt Cloud, which compete with Soda Core/Cloud for the data transformation and quality layer of the modern data stack.
Emerging players
- Datafold: Datafold is an emerging player focused on data diffing, regression testing, and data observability for dbt and warehouse workflows; it overlaps with Soda's pipeline testing and data quality monitoring use cases for engineering-led teams.
- Anomalo: Anomalo is an emerging player in automated data quality monitoring for warehouse tables with no-code configuration; it competes for the same data steward and data engineer personas as Soda Data Observability.
- Atlan: Atlan is an emerging player in the data catalog and governance space that bi-directionally integrates with Soda to surface quality results; it competes for the same governance and data steward buyer while partnering on quality execution.
Broad incumbents
- Datadog: Datadog acquired Databand in 2022 to enter data observability, and offers broader infrastructure and APM monitoring with overlapping data quality and pipeline monitoring capabilities that compete with Soda for enterprise platform budgets.
- Collibra: Collibra is a broad incumbent in data governance and cataloging with bi-directional integration with Soda, but also competes indirectly by bundling data quality rules and stewardship workflows into its enterprise platform.
- Informatica: Informatica is a broad incumbent offering enterprise data integration, data quality, and governance as part of a large platform; it competes with Soda in regulated enterprises where Informatica bundles data quality into its broader data management stack.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Customer concentration
Soda social profiles
Digital presenceSoda compliance and trust
Trust signalCompliance4 records
Soda financial estimates
Financial estimateRevenue estimate
Valuation estimate
Soda leadership team
Management profileNumber of profiles
Profiles6 records
Soda subsidiaries and ownership
Company hierarchySubsidiaries1 record
Soda funding detail
Funding detailFunding overview
Funding rounds4 records
Investors6 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Soda M&A and investment
M&A and investmentM&A1 record
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Soda
What does Soda do?
Soda provides an AI-native data quality platform that detects, explains, and resolves data issues across enterprise data pipelines. The platform combines Soda Core (open-source YAML/SodaCL testing engine) with Soda Cloud (hosted SaaS control plane) and Soda AI (proprietary AI capabilities), delivering data observability, collaborative data contracts, record-level anomaly detection, smart alerting, a Diagnostics Warehouse, and agentic data cleansing. The offering is sold as free, Team ($750), and custom Enterprise subscriptions, with usage-based Soda Processing Units (SPUs) and an add-on for Soda Cleanse agentic remediation.
Is Soda a public or private company?
Soda is a private company. It is classified as venture growth investor backed and is currently operating.
When was Soda founded?
Soda was founded in 2018. It employs 11 to 50 people.
Where is Soda based?
Soda is headquartered in Brussels, Belgium, in the Europe region.
How does Soda make money?
Five revenue lines are on record. Freemium tier is the primary driver. The others are team subscription, enterprise subscription, soda Processing Units (SPU) overage and soda Cleanse add-on.
Who are Soda's main competitors?
Direct peers on record are Monte Carlo Data, Bigeye, Great Expectations and dbt Labs. Emerging players are Datafold, Anomalo and Atlan. Broad incumbents are Datadog, Collibra and Informatica.
Does Soda have an API?
Yes. Soda exposes a programmatic API, a CLI, and an MCP (Model Context Protocol) endpoint that allow engineers to drive Soda through code, automate data quality workflows across pipelines, and integrate with agents such as Claude, Codex, and Cursor. Contracts and quality status live in the UI and over MCP, CLI, and API so engineers, stewards, and AI agents work from the same definition of good data. Documentation is published at docs.soda.io. The MCP endpoint for the Soda Cleanse remediation workflow is on the roadmap. Developer documentation is at docs.soda.io.