Titulo Estágio
Scalable Real-Time Metrics
Áreas de especialidade
Engenharia de Software
Engenharia de Software
Local do Estágio
Coimbra OU Lisboa OU Porto
Enquadramento
About Feedzai:
Feedzai is a company that makes bleeding-edge machine learning software.
The world’s mightiest payment networks, banks and retailers use us to prevent fraud when customers shop in store, online or via mobile devices. Backed by years of hardcore work and funding from amazing investors (Oak HC/FT, Sapphire Ventures, Data Collective) we’re at the inflection point of growth.
Context:
Feedzai’s fraud detection platform consists on a machine learning platform with an integrated runtime for real-time fraud detection on electronic payments.
One of the key requirements of this platform is the processing of real-time transactions streams with a high volume of transactions per second and take a decision regarding its fraud status in less
than 3 milliseconds. With such aggressive time constraints, it means that one of the core engineering goals of Feedzai is to guarantee that our technology stack used in the entire platform fits our needs while guaranteeing scalability, fault-tolerance, high throughput and low latency.
Some of the technologies integrated in the platform are: Apache Cassandra, RabbitMQ, Apache Spark and Zookeeper.
To ensure the required low latency numbers, at the core of this platform there’s a high-performance Event Stream Processing engine that allows us to compute thousands of real-time metrics per second with millisecond latencies. The current solution is based on an in-memory engine developed in-house, which even though makes it possible to comply to such low latency requirements, adds statefulness requirements to the product, which in turn poses big challenges
for scalability.
Objetivo
Objectives:
The goal of this internship is to discover and implement a novelty way to compute real-time metrics in a scalable, distributed and fault-tolerant manner, while at the same guaranteeing the
low-latency performance that is required of Feedzai’s platform.
You’ll have the chance to work with exciting technologies and have a huge impact on the way that Feedzai detects fraud.
To achieve the objectives, the candidate must study the state of the art on distributed metric computation engines, identify the key characteristics in terms of workload for each use-case, and
set up a methodology which replicates each workload and benchmark at least one technology from each type of distributed engine.
After the best engine is selected, a Proof of Concept should be developed and integrated within Feedzai’s platform.
Plano de Trabalhos - Semestre 1
Work Plan:
Get familiar with Feedzai’s platform and Feedzai’s business
Study the state of the art on distributed computation engines
Define the methodology and the use-cases that will be used to benchmark the solution
Plano de Trabalhos - Semestre 2
Work Plan:
Proof of Concept and benchmarking for at least three engines and write intermediate
report about findings
Implement and integrate the chosen technology within Feedzai’s codebase
Write the final report
Condições
PC
Horário de trabalho flexivel
Bolsa de Estágio
Possibilidade de realização do estágio no escritório de Coimbra, Porto ou Lisboa.
Orientador
Ricardo Ferreira
ricardo.ferreira@feedzai.com 📩