Pedro Marques, Developer in Zürich, Switzerland
Pedro is available for hire
Hire Pedro

Pedro Marques

Verified Expert  in Engineering

Software Developer

Zürich, Switzerland

Toptal member since November 5, 2021

Expertise
Machine LearningC++NLPSaaSLinuxPythonDockerGolangKubernetesTensorFlowGoogle Cloud Development
Bio

Pedro is an experienced software developer, team lead, and architect who has built thriving products in embedded systems, cloud platforms, SaaS applications, and machine learning. He has startup experience both as a company founder of one successful exit and senior technical staff. He led Fortune 500 companies' technical projects for Juniper Networks and Google. For the past five years, Pedro's focused on machine learning, leading projects at Laserlike (acquired by Apple), Google, and startups.

Portfolio

BLP Digital
TensorFlow, Python, Keras, Natural Language Processing (NLP)...
Google
C++, Natural Language Processing (NLP)...
Laserlike
Go, Google Bigtable, Google Cloud, Kubernetes, Data Science

Experience

  • C++ - 10 years
  • Python - 8 years
  • Generative Pre-trained Transformers (GPT) - 5 years
  • Natural Language Processing (NLP) - 5 years
  • Go - 4 years
  • Kubernetes - 4 years
  • SaaS Design - 4 years
  • Keras - 3 years

Availability

Part-time

Preferred Environment

TensorFlow, Keras, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), C++, Python, Go

The most amazing...

...product I've built is a network virtualization solution called Contrail Networking, which is deployed by some of the largest telecom companies in the world.

Work Experience

Founder and CTO

2019 - 2020
BLP Digital
  • Designed the application structure, including the SaaS platform. I created the engineering processes and frameworks for both ML and cloud software delivery.
  • Built a TensorFlow model to interpret tables in commercial documents such as invoices and delivery notes using a combination of NLP and image processing techniques.
  • Created the application back end that interfaces between the user and the ML predictions.
Technologies: TensorFlow, Python, Keras, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT)

Tech Lead

2017 - 2019
Google
  • Led a project to extract entity relationships from web documents.
  • Created a data extraction pipeline over Google's document indexing.
  • Built an integration between extracted tuples and the Google search engine.
Technologies: C++, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT)

Senior Software Enginer

2016 - 2017
Laserlike
  • Developed a machine-learning algorithm to embed web documents in a topic space. Created a search algorithm to retrieve documents by topics.
  • Built a GitOps automation system to deploy the application stack to Google Cloud.
  • Built a data processing pipeline to extract data from Bigtable, clean up and process the dataset, and train the document embedding models.
Technologies: Go, Google Bigtable, Google Cloud, Kubernetes, Data Science

Founder and Lead Architect

2012 - 2016
Contrail Systems
  • Created the product concept, recruited the founding team, and participated in the process of raising capital from investors.
  • Built the original prototype of the control plane.
  • Supported the initial lead customers and through their deployments.
Technologies: C++, Kubernetes, OpenStack, Border Gateway Protocol (BGP), Amazon Virtual Private Cloud (VPC)

Experience

TensorFlow and Keras Contributer

https://github.com/pedro-r-marques/keras-list-mapper
As part of an intelligent document processing model, I built a Keras model using RaggedTensors to deal with varying document sizes. This required building generic libraries to help handle variable-size documents and bug fixes to the TensorFlow and Keras codebase.

Linux Kernel IPv6 Stack

I wrote the original implementation of the Linux Kernel IPv6 stack (1996) and helped maintain the TCP implementation in the kernel. Added TCP functionality such as fast retransmit and improved congestion control mechanisms.

Kubernetes Contributor

I handled bug fixes and contributed to the Kubernetes node networking stack to add support on Kubernetes for additional pod networking solutions. I also built an integration between the Contrail Networking product and Kubernetes and OpenShift.

Education

1990 - 1995

Bachelor's Degree in Computer Science

Universidade de Lisboa - Lisbon, Portugal

Skills

Libraries/APIs

Keras, TensorFlow, Pandas, Scikit-learn, SpaCy

Tools

Amazon Virtual Private Cloud (VPC), Flume

Languages

C++, Python, Go, SQL

Frameworks

LightGBM

Platforms

Kubernetes, OpenStack, Docker, OpenShift, Amazon Web Services (AWS)

Storage

Google Cloud, Google Bigtable, MySQL

Other

Border Gateway Protocol (BGP), Machine Learning, Natural Language Processing (NLP), Computer Science, SaaS Design, Linux Kernel Programming, Transmission Control Protocol (TCP), IPv6, Data Science, Generative Pre-trained Transformers (GPT)

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring