$$$
{{ $t($store.state.user.experience_value_in_dollars) }}
Senior
{{ $t($store.state.user.experience_search_name) }}
0
jobs
Machine Learning Research Engineer
Farouq Oyebiyi
,
Leeds, United Kingdom
Experience
Other titles
Skills
I'm offering
Applied Researcher with experience developing compute-efficient Machine Learning algorithms to extract information from Images, Text and large databases. Specifically, Object Detection, Image Segmentation, Information Retrieval and Sequence Tagging.
Markets
United Kingdom
Language
English
Fluently
Ready for
Larger project
Ongoing relation / part-time
Available
My experience
2019 - ?
job
Research Engineer
Leeds Beckett University.
• Climate Change
o Conducting research on using Machine Learning to estimate Embodied Carbon of infrastructures from associated data such as blueprints, floor plans and design
model metadata.
• Quality Control
o Built a sequence tagger to extract information from highly unstructured non-
conformance reports. Fed extracted information to a clustering model to discover
factors contributing to low quality of delivered infrastructure.
o Conducting research on using Machine Learning to estimate Embodied Carbon of infrastructures from associated data such as blueprints, floor plans and design
model metadata.
• Quality Control
o Built a sequence tagger to extract information from highly unstructured non-
conformance reports. Fed extracted information to a clustering model to discover
factors contributing to low quality of delivered infrastructure.
Design, Machine learning, Research, Infrastructure
2018 - 2019
job
Lead Machine Learning Engineer
InstaDeep.
• Optical Character Recognition (OCR)
o Led a team of Machine Learning Engineers in researching literature and implementing state-of-the-art Optical and Handwritten Character Recognition,
achieving Character Error Rate of ~ 3.5%.
o Led the Artificial Intelligence team consisting of 5 direct reports, all of whom I
hired and mentored.
o Led a team of Machine Learning Engineers in researching literature and implementing state-of-the-art Optical and Handwritten Character Recognition,
achieving Character Error Rate of ~ 3.5%.
o Led the Artificial Intelligence team consisting of 5 direct reports, all of whom I
hired and mentored.
Machine learning, Artificial Intelligence, OCR, LED
2017 - 2018
job
Data Scientist
Criterion Analytics.
• Object Detection and Image Segmentation
o Built a model off pretrained Mask R-CNN in PyTorch to detect and segment
clothing items.
o Developed filters using OpenCV for post-processing model predictions.
o Researched and built prototypes of efficient anchor box generation for object
detection model.
• Information Extraction for low-resource language
o Tracked state-of-the-art research in sequence tagging and implemented models
using TensorFlow.
o Built a Bi-LSTM-CRF parser with word embeddings as input to extract named
entities from legal documents. Augmented this with hand-written rules to push
accuracy to 95%.
• Learning to Rank
o Built a ranking model using Gradient Boosted Trees for a service marketplace such
that services are matched to the vendors with the highest likelihood of completing
tasks on time and without complaints. This had a 20% increase on the existing
method.
• Credit Scoring
o Used alternative data sources such as rent and utility payments to build a Regression
model that estimates the credit score of users.
o Built an automated ETL pipeline based on a 5-node Apache Spark cluster in Google Cloud for processing large datasets (~1TB).
o Educated team on importance of not using discriminatory factors such as race and gender when building models.
o Led the project, from ideation, up till deployment and production monitoring. Co-
ordinated the efforts of 2 direct reports, 2 data vendors and 1 indirect report.
o Built a model off pretrained Mask R-CNN in PyTorch to detect and segment
clothing items.
o Developed filters using OpenCV for post-processing model predictions.
o Researched and built prototypes of efficient anchor box generation for object
detection model.
• Information Extraction for low-resource language
o Tracked state-of-the-art research in sequence tagging and implemented models
using TensorFlow.
o Built a Bi-LSTM-CRF parser with word embeddings as input to extract named
entities from legal documents. Augmented this with hand-written rules to push
accuracy to 95%.
• Learning to Rank
o Built a ranking model using Gradient Boosted Trees for a service marketplace such
that services are matched to the vendors with the highest likelihood of completing
tasks on time and without complaints. This had a 20% increase on the existing
method.
• Credit Scoring
o Used alternative data sources such as rent and utility payments to build a Regression
model that estimates the credit score of users.
o Built an automated ETL pipeline based on a 5-node Apache Spark cluster in Google Cloud for processing large datasets (~1TB).
o Educated team on importance of not using discriminatory factors such as race and gender when building models.
o Led the project, from ideation, up till deployment and production monitoring. Co-
ordinated the efforts of 2 direct reports, 2 data vendors and 1 indirect report.
Spark, UP, Object Detection, LED, Processing, Production, BEE, Google, OpenCV, Apache spark, Monitoring, Research, Service, Node, Tensorflow, Google cloud, Apache, Word, Cloud, ETL, R, Deployment
2014 - 2017
job
Software Engineer, Machine Learning
Konga Online Shopping.
• Recommender System
o Pitched creation of in-house Machine Learning expertise to management. Helped
set up a team and drove first project from ideation till deployment.
o Led a team of engineers to build and deploy to production a large scale ETL
pipeline using Apache Spark, Apache HBase and Elasticsearch.
o Developed and deployed recommender services using explicit and implicit
feedback data to match customers to products they are most likely to buy. Models
were a hybrid of Topic Modeling and Collaborative Filtering to combat cold-start.
Recommender services included:
§ People Also Bought
§ Based on Purchase History
o Ran A/B tests to determine how recommender models influenced customers'
purchase decisions.
o Pitched creation of in-house Machine Learning expertise to management. Helped
set up a team and drove first project from ideation till deployment.
o Led a team of engineers to build and deploy to production a large scale ETL
pipeline using Apache Spark, Apache HBase and Elasticsearch.
o Developed and deployed recommender services using explicit and implicit
feedback data to match customers to products they are most likely to buy. Models
were a hybrid of Topic Modeling and Collaborative Filtering to combat cold-start.
Recommender services included:
§ People Also Bought
§ Based on Purchase History
o Ran A/B tests to determine how recommender models influenced customers'
purchase decisions.
Machine learning, Deployment, ETL, Elasticsearch, Apache, Management, Spark, Apache spark, Software, Production, Hybrid, LED, UP
My education
2018
-
2021
Georgia Institute of Technology
Masters (Part-time), Data Science
Masters (Part-time), Data Science
2009
-
2013
Fountain University Osogbo
BSc, Computer Science
BSc, Computer Science
Farouq's reviews
Farouq has not received any reviews on Worksome.
Contact Farouq Oyebiyi
Worksome removes the expensive intermediaries and gives you direct contact with relevant talent.
Create a login and get the opportunity to write to Farouq directly in Worksome.
38000+ qualified freelancers
are ready to help you
Tell us what you need help with
and get specific bids from skilled talent in Denmark