We help you create and scale your data architecture from scratch

Having the right data and having it on time is a game changing factor in businesses. Our data engineering teams have helped several banks, insurance companies build up their data architecture leveraging cutting edge cloud and on premise technologies. Our attention to performance and data quality has been pivotal in getting the implementation right from the first time.

Technologies
We chose the best technologies
Putting together a data infrastructure isn't a difficult task as many tools offer that possibility. Putting together a strong data infrastructure that scales without losing performance is a whole different story. That is why our expertise in data engineering and Big Data has been a remarkable asset for our clients. We carefully chose our technology stack and we took the time to master it.
Python
Python
Python is a very popular high-level language that is renowned for the readability of its code. Instagram, Spotify and Amazon have chosen Python. The ecosystem of Python and the large number of frameworks allows the front and back ends of the project work better together.
Airflow
Airflow
Airflow is a workflow management platform. It started at Airbnb as a solution to manage the company's increasing complex workflows. Our developers like Airflow because it allows them to configure retry policies into individual tasks. It also allows them to set up alerting in the case of failures, retries, as well as tasks running longer than expected.
Spark
Spark
Spark is a distributed data processing engine that is suitable for use in a wide range of circumstances. Spark allows our developers to rapidly query, analyze, and transform data. Our developers use it to launch large data sets, processing of streaming data from sensors, IoT, or financial systems, and machine learning tasks.
Hadoop
Hadoop
Hadoop is a framework that allows for the distributed processing of massive data sets across clusters of computers using simple programming models. Rather than rely on hardware to deliver high-availability, Hadoop itself is designed to detect and handle failures at the application layer. A wide variety of companies and organizations use Hadoop for both research and production due it is reliable and extremely scalable.
Case Study
Design and implementation of a Smart search engine

A universal bank present in 32 countries in Asia, Europe & Africa with more than 14,000 employees and around 500 millions euros of net banking income.

Several thousand of employees operating in potentially complex areas especially foreign trade More than 30,000 various informative documents (Word, PDF, PPT, Excel, web pages) thanks to massive investment in documentation (international regulations, exchange, product sheets, etc.) Strong difficulty for employees to navigate through this knowledge base despite the investment in a search tool (Dassault Systems - Exalead)

Objective

Design and implement a powerful efficient search engine covering the bank's large document and knowledge base.

Search engine has to provide good results whatever the formulation of the research or at least give great flexibility on the formulation.

It also must provide comfort in research and allowing easy access to the information the user is seeking

Results

Fully agile methodology project with mixed team : Theodo & bank.

Tool similar to an internal "Google" and consists of 2 complementary modules :

Graph search module

Interpretation of the object of research by a Theodo proprietary semantic graph rich in financial vocabulary developed by indexing several banking and financial institutions public material

Learning Module

Collecting implicit user feedback to improve the relevance of the results (including associated searches, etc.)

Flexible and efficient search engine made available in production to thousands of employees.

Search engine implemented proved to be far superior than the pre existing search engine (Exalead) on a sample of more than 50 critical search use cases.

Far better user experience

- User can preview and navigate answer before downloading any documents

- Relevant answer highlighted to facilitate reading

Technologies:
Case Study
RPA engine to automate accounting processes

A multinational insurance player with more than 930 employees & more than 580 millions USD turnover.

The insurance and reinsurance company is the parent company and has equity interests in a multitude of activities Tedious consolidation work and time-consuming for the accounting department (several days) Identification of 2 high impact use cases with immediate value

Objective

Implement 2 RPA use cases to free up time for accounting agents and optimize processing times :

- Automation and synthesis of raw supply and unpaid invoices data stemming from +50 excel sources Consolidation of accounting data by subsidiary / participation given these subsidiaries / participations have different approximative names in different sources (fuzzy matching necessary)

Results

Project with mixed team : Deployment of the Theodo team and Workshops & design sprints with accounting department

2 bespoke designed and implemented RPA engines

Thanks to the RPA deployed in production, processing time by the accounting agents came down from 10-15 man-days per quarter to 0 (algorithm takes 15min to execute without any manual intervention).

We did this in 6 weeks of intervention

Technologies:
Read more about Data Engineering expertise
View all articles
Launch your digital project with us today! 🚀Get in touch with someone who will quickly understand your need and link you directly to our experts!Start my project