Realtime Data Connectors
MovingLake is Fivetran for event-driven architectures. Companies such as Casai use our product to obtain orders and price changes in real time.
Active Founders
Andres Paez
Worked as a Software Engineer and Product Manager at Amazon Prime, Google Search and Google Cloud Platform for about 7 years. Have worked as the CTO of Casai and Valoreo for the last 3 years starting technology teams from scratch leading software engineers, hardware engineers, software managers, data engineers, data scientists, IT professionals, DevOps engineers, product managers and product designers.
Edgar Cabrera
About ten years of Sofware Engineer Experience at startups of several stages. I have worked in developing products from scratch until they become scalable, robust systems. I love the feeling of my work bringing value to others.
Company Launches
Hey YC!
Edgar Cabrera and I are excited to launch Moving Lake today.
TLDR: MovingLake aims to do ETLs for event-driven architectures. Power your data warehouse as well as your microservices using the same API connectors.
⚠️ What’s the Problem?
After talking to dozens of data engineers as well as technology stakeholders, and from our past experience, we uncovered a set of intertwined problems with current data connectors:
- A data connector does not provide data to transactional systems. This can lead to duplication of integration work plus lack of reliability between different sources of data. For machine learning this can be real trouble
- API integrations done in house are flaky and require a lot of investment in development and maintenance
- Event-driven architectures are reimplementing over and over again the same pieces of infrastructure such as schedulers, pollers and webhook receivers
- Batch extractors can skip states and so end up with missing data. For example if I have a booking business and a reservation gets cancelled 10 minutes after created, the batch pipeline will only see the cancelled state but not the created state.
- Running automations and other workflows is impossible with batch data connectors.
👍 What’s the solution?
Realtime, event-driven, deduplicated, replayable, ELT friendly and SDK ready connectors which feed both data and transactional systems at the same time.
We are rolling out connector for APIs which give you the ability to add any number of destinations including data and backend systems. We provide on top of that transformations for the data side so that you get the data in nicely formatted tables. Plus we also aim to provide easy to use SDKs so that consuming MovingLake’s webhooks is as easy and fast as possible.
❓ What’s the ask?
Do you have recurring issues with data reliability? Do you end up writing batch scripts to recover lost data from webhooks? Do you struggle with running your machine learning models from batch pipeline’s data? Reach out to founders@movinglake.com or contact us through our website.
Thanks!