Fast, Reliable Data Access Helps Phenix Fight Food Waste

Key Takeaways

• Saved six months on manual pipeline builds
• Less than one hour a week on in-house maintenance
• 30 million lines of data generated each month
• Fivetran dbt transformations offered hosting of open-source modelling
• Saved the equivalent of half the salary of a full-time data engineer

Data Stack

• Pipeline: Fivetran
• Source: Salesforce, NetSuite, Intercom, Open Food Facts
• Destination: Google BigQuery
• Business Intelligence: Looker

Phenix is a “tech for good” company on a social and environmental mission to redistribute 10 million tons of food wasted in France each year to eight million people affected by food scarcity.

The Challenge: New Architecture Needed to Access Quality Data

Recognising that Fivetran could accelerate the process of redistributing food from supermarkets, producers, manufacturers, and wholesalers to the people who need it, Phenix launched a web platform where unsold food products could be made available at reduced prices. The next challenge was to find an effective way to turn data captured on the website and in a range of applications into actionable insights.

The company knew a deeper understanding of interactions with customers, associations, and the food industry would not only help its work in the field, but also make the company more efficient at marketing and forging stakeholder relationships.

An architecture was needed to set up data flows from databases to a data lake and data warehouse, and there had to be a way to ensure that data collected upstream would be clean and usable downstream.

The Solution: Key Metrics to Increase Operational Efficiency

As well as the web platform, Phenix had a range of data-rich applications that it needed to access: Salesforce for CRM, NetSuite for financials and Intercom for customer support. Public data sources like Open Food Facts would also need to be connected.

Fivetran was chosen to provide connectors for four main reasons:

1. It is scalable and could easily handle the 30 million lines of data generated each month.
2. It could provide prebuilt connectors to the key applications and sync all the data from the different source databases
3. The support, alerting and monitoring system that is integral to its managed service avoids pipeline breaks that can disrupt the flow of data.  
4. The Fivetran-dbt transformation offering includes hosting the open-source modelling tool. Phenix uses dbt for pipeline tests and checking for duplicates, speeding up the process of bringing clean and accurate data to dashboards in Looker, the company’s BI tool.

Andy Barakat, Data Engineer at Phenix, summed up the success of the project: “Fivetran has enabled us to consolidate and centralise all the data sources in a data warehouse.” The result is data modelling that delivers key metrics to help make Phenix more effective, such as re-valuations of unsold products, rates of breakage that leads to waste, and the number of tons of products saved and equivalent impact on CO2 emissions.

The Results: Managed Service Saves Time and Money

Phenix saved six months from the immediate adoption of Fivetran, eliminating the time it would take trying to develop and build connectors in-house. And because the pipelines are provided as a managed service, less than an hour a week is spent maintaining them internally. As Andy Barakat puts it:

We don't have to spend any time updating the APIs. Whenever the source changes or a new field or column is introduced, it automatically changes in our own database. Fivetran manages it all for us. It runs in the background, which means we can focus on the transformation and modelling of the data — areas where we really add value.

Another objective that Fivetran helped achieve was enabling all teams to have access to data via weekly reports that they use to make strategic business decisions. The company plans to become more data-driven have been accelerated.

The Future

The next stage is to fine-tune recommendations and use algorithms to build models with historical data that will reduce food waste in the future. With GDPR standards built into the product, Fivetran has also played a role in meeting regulatory requirements, saving time by helping Phenix stay compliant around sensitive data.

Phenix has calculated that Fivetran has saved the equivalent of half the salary of a full-time data engineer. Andy Barakat states:

Fivetran helped us save a tremendous amount of time and money, we don’t have to reinvent the wheel. They do the connectors and ELT very well.

About Fivetran: Fivetran, the leader in automated data integration, delivers ready-to-use connectors that automatically adapt as schemas and APIs change, ensuring consistent, reliable access to data. Fivetran improves the accuracy of data-driven decisions by continuously synchronizing data from source applications to any destination, allowing analysts to work with the freshest possible data. To accelerate analytics, Fivetran automates in-warehouse transformations and programmatically manages ready-to-query schemas. Fivetran is headquartered in Oakland, California, with offices around the globe. For more information, visit fivetran.com.‍