Every data governance program lacks one and only one thing: Execution. Data stewards understand the standards but are never the ones who can enforce them. They know what "correct" means for your data better than anyone: ➤ order_id should never be null ➤ orders should land by 6am ➤ currency should be one of a fixed set of values ➤ revenue should never drop more than 20% day over day But knowledge and execution are two different things. To enforce any of it, someone has to write checks or tests, configure the monitor, wire it into the pipeline as a CI/CD gate, and deploy it. MCP unlocks this capability for all data management tasks. You can now connect an AI assistant to your pipelines and describe the various rules in plain language. The assistant maps it to the right tools - drafts the contract, configures the monitor, runs the check - and you can review and approve before anything deploys. Interested in getting started? Hakim is running the full data quality lifecycle live through MCP. He will explain how you can: ➤ Connect an AI assistant to your data quality tooling with MCP — and what that unlocks ➤ Go from an unmonitored dataset to an enforced data contract in minutes, through conversation ➤ Scale contract coverage across your estate Register today!
Soda
Software Development
Brussels, Brussels 24,484 followers
The AI-native, fully automated data quality platform. From detection to resolution.
About us
Most companies struggle to operationalize data governance and quality. Business teams don’t want to manually enforce rules, and engineers get buried in pipeline issues — eroding trust in data and slowing innovation. Soda fixes this with the only end-to-end data quality platform that automates the entire workflow — from detection to resolution — with AI built for data quality. We meet users where they are: - Engineers manage everything as code in Git. - Business users create and review data contracts in a collaborative interface. - Together, they work in a shared, AI-powered workflow to define quality expectations, monitor metrics, and isolate and remediate bad data directly in their environment. By uniting teams, automating with AI, and securing trust at the source, Soda helps organizations like Disney, Nubank, and HelloFresh restore confidence in their data and decisions. Why Soda? - Best AI for Data Quality — purpose-built, faster, and more accurate, with 70% fewer false positives than traditional monitoring. - Unite Business and Engineering — collaborative data contracts that bridge governance and technical workflows. - Securely Isolate and Fix Bad Data — record-level anomaly detection and remediation inside your own environment. Soda brings width and depth to data quality — from every dataset across multiple warehouses to every individual record in a dataset. Join us in building a world where teams trust their data, decisions, and AI.
- Website
-
https://soda.io/
External link for Soda
- Industry
- Software Development
- Company size
- 11-50 employees
- Headquarters
- Brussels, Brussels
- Type
- Privately Held
- Founded
- 2018
- Specialties
- data quality, data management, data science, data engineering, data monitoring, data observability, data testing, data reliability, data product management, data analytics, data products, data mesh, data, and data pipeline management
Products
![Click here to view Soda Data Quality]()
Soda Data Quality
Data Quality Software
Soda is the only end-to-end data quality platform that automates the entire workflow — from detection to resolution — with AI built for data quality. It unites business and engineering teams in a shared, collaborative workflow to define, monitor, and enforce data quality standards. With Soda, engineers manage checks and contracts as code in Git, while business users review and approve them in an intuitive interface. Our proprietary AI detects anomalies with record-level precision, automates quality checks, and even isolates and remediates bad data directly in your environment. Trusted by global leaders like Disney, Nubank, and HelloFresh, Soda delivers scalable, secure, and explainable data quality — so teams can trust their data and the decisions they make with it.
Employees at Soda
View 129 employees at Soda
or
Locations
-
Primary
Get directions
Brussels, Brussels 1000, BE
-
Get directions
New York, NY, US
Updates
-
As AI becomes part of everyday data workflows, data quality can no longer be an afterthought. Building reliable pipelines requires more than checks—it requires a workflow that can scale, adapt, and be understood by both humans and AI. Join Santiago for a live session on how Soda CLI helps teams build modern data quality workflows. Using a real warehouse, he'll demonstrate how to move from schema discovery to data contracts, anomaly detection, and AI-assisted execution in a single workflow. This session is for data engineers, analytics engineers, and anyone looking to get ahead with AI-powered data workflows. In this webinar, Santiago will show you how to: ➤Connect a data warehouse and discover datasets from the Soda CLI ➤Generate a data contract directly from a live schema ➤Add anomaly detection monitors, including both column-level monitors and custom SQL checks ➤Hand the same workflow to an AI agent and watch it execute the steps autonomously Whether you're just getting started with Soda CLI or want to see how it fits into AI-assisted data quality workflows, this session will give you a practical, end-to-end demonstration.
-
Have you been hearing the term “data contracts” everywhere, but aren’t sure how to actually implement them in your pipelines? Data contracts are becoming standard in modern data architectures, yet most teams still struggle to move from concept to execution — especially when governance and engineering work in different tools and languages. In this webinar, you’ll walk through a practical, step-by-step approach to implementing data contracts in a way that works for both technical and non-technical roles. You’ll learn how to: ➤ Understand and design the key elements of an effective data contract ➤ Work with data contracts in code-first (YAML) for engineering workflows ➤ Manage and maintain the same contract using a no-code interface This webinar is for you if you are: ➤ A data engineer or analytics engineer who needs contracts in CI/CD and pipelines ➤ Someone responsible for data quality, ownership, or compliance across datasets
-
Soda reposted this
Your Most Extroverted Geek... Never thought it'd be my entire team name! 😅 Last Saturday, Hackathon! Though NOTHING went as expected My expectations? A team of 5, EdTech clear use case, company data of 80k users. Me? The sales guy, getting pulled by my technical team. 😅 The AI guy? Sick. Together with the security guy... Our Tech lead and UI/UX Artist had lost hope, they stayed behind. I stepped off the train in Ghent, alone. I was resolved, going I must! 😅 Got my first Mac last week. No npm, 0 MCP, not even GitHub connected. From ground 0, and logging it all in took me the whole morning. 😅 No access to company data, no idea what I'd be building. I was ready to complement, not to invent groundbreaking. Took a deep breath. And three coffees. I hesitated joining a team, though after an hour it was pretty clear the Loop Earplugs track was where my heart was. Out of three tracks, the sole with a scope. Marketing Strategy, Data Analytics, topics I felt attracted to. 🥳 Your Bruxellois is hardly alone! Alexandria Bohn and Patricio Castillo welcoming me right off the bat was a life saver! You both gave me hope, inspiration, and encouraged my quirks. Thank you. And to then see Bart BECKS and Stijn (Stan) Christiaens to close off, absolutely legendary! I cherish our talks. Let's make Brussels ROCK! 🥳 Claude hard-carried the install of all packages I needed, the GitHub connections, etc. Setup = Raycast, CapCut, Warp CLI + Claude, Soda and GitHub connected inside. + some more Claude next to it for good measure. It was my first CLI-agentic experience and a blast! I recommend it. 🥳 And above all: Thank you Loop Earplugs for access to immaculate real-life data! As a serial early-stage enthusiast, I had never played with numbers of a 100-employee firm, and I can confidently say I loved it! No open repo, as data came with an NDA. Two focus points to share though : - Making submission fit within Loop Earplugs own structure. Had read the challenge 10 times. I won on following a script. - Embedding the challenge in a marketing campaign. Between Claude prompts, I was shooting videos. Dancing at the DJ booth or napping between tables. I broke the fourth wall by being the actor of the brief I designed. Thank you Maksym Dziarmaga 🧲 for the on-stage energy! Thank you for the airplugs. And for the sizeable prize 👀 If you did not attend: 👇🏼 Guess what the prize is in the comment 👇🏼 Now back to work - AKA.- Organising our own hackathons 🤣 13th June - Energy grid with Ville de Bruxelles - Stad Brussel and Sibelga 27th June - Agentic Commerce II with Stripe Community, Collibra and e-invoice.be 👇🏼 Drop HACKATHON to receive an invite! 👇🏼 Thank you to Wintercircus Ghent, Soda, Anton Wouters, Wiljan Cools for organising a fabulous Hackathon! And to the STACKED lineup of sponsors: Anthropic, Google, Aikido Security, Lovable, Cursor, Tinrate North Star, Manual.to, Happy Client, Syndicate One, Scaleup Flanders, Red Bull, Troom, Adventure Venues
-
-
Soda reposted this
Work that would have taken me days, done in a few hours. Last weekend I took part in the Hackers & Ravers hackathon at Wintercircus Ghent, organised by Soda. I teamed up with Kwan Jau Siu, a seasoned accounting professional who brought a problem straight from his field: procurement data is messy. When purchase orders are typed manually into an ERP, errors slip through — and tracking them down afterwards is slow and frustrating. So we built a Purchase Order validator: a demo that catches those errors at entry and shows how much of the manual cleanup can be avoided. Two things stood out for me. First, the chance to work on a genuine problem rather than a contrived one — it kept us focused. Second, the speed: using Cursor with Claude Opus 4.8 from Anthropic, we compressed into a few hours work that would have taken me days just a few months ago. It's a demo, of course — but a working one. Thanks to Soda and Wintercircus Ghent for putting together an event that makes this kind of building possible. Repo in the comments 👇
-
-
Soda reposted this
Last weekend Saturday (30th of May) I dedicated my weekend for a very special event “Hackers and Ravers”. It was organised by Soda + Wintercircus Ghent. This HACKATHON gave the attendees possibility to compete in mainly 2 different categories. As our team was attracted to the prize pool of mac minis and other additional prizes we chose the BUILDERS BATTLE category. Long story very short, WE DIDNT WIN but thanks to this event I realised where the current AI development is and what it is capable of (tbh this capability I believe is just the tip of an iceberg). More about what our team offered to this HACKATHON is an agentic AI structural feature where we gave the possibility to an AI to create agents, and these newly created agents have also the possibilty to become parent to other agents (creating a tree like structure). Such structural feature could be imo be useful for complex problems where they can be divided into sub problems and solved in smaller chunks. Lastly, thanks for my team mates, organisers and sponsors for their hard work. Hope it will be organised again next year!
-
-
Soda reposted this
Had a great time this saturday at Hackers & Ravers. 12 hours of AI driven coding, from idea to product. Together with Keano I represented Dynamate. We built a small application that connects our company's resources (Drive, Slack, Teamwork,...). Istead of searching multiple platforms yourself, you ask one question and an agent pulls everything together. Thanks to Soda and Wintercircus Ghent for organizing!
-
-
Soda reposted this
I won the builders track at the Hackers and Ravers Hackathon! A HUGE thank you to the incredible people at Soda and Wintercircus Ghent for organizing such a wonderful event. Special shoutout to Anton Wouters for all the effort! I had a great time building what I think the future of interfaces could look like. You can check out the project here: https://lnkd.in/eS63eAj2 It was built in under 6 hours, so please cut me some slack. More than the prize and the lovely goodies, I’m grateful that I got to meet some amazing people and yap about things I usually don’t get to. Also, a big thank you to Aikido Security, Loop Earplugs, Anthropic, Google, Tout Bien Pils 🥳 , and all the other supporters of the event for making it such a unique and high-quality experience.
-