In the Code: Volume 1, Issue 4
In Issue 4 of our LinkedIn Engineering In the Code e-newsletter, we feature an arXiv paper via our Product Strategy AI VP Xavi Amatriain and our Engineering team on the Transformers class of deep-learning models.
We also highlight additional LinkedIn Engineering career stories, technical work, news stories, and some of our latest job openings.
The Transformers class of deep-learning models has been highly effective for a variety of Natural Language Processing (NLP) tasks, with the advent of dozens of new models in recent years.
To better catalog and classify these, LinkedIn Engineering’s Product Strategy AI VP Xavi Amatriain and others from LinkedIn Engineering's AI, Machine Learning (ML), and Systems & Infrastructure teams launched an arXiv paper here.
Their catalog includes models trained via self-supervised learning (e.g., GPT3, etc.) and human-in-the-loop.
Career stories: Diving into software engineering at LinkedIn
Initially a Chicago-based data analyst, Jelanah had her heart set on a more meaningful career in frontend (UI) engineering.
She chronicles for us how a life-changing engineering mentor at LinkedIn unlocked her frontend technical skills, supported a relocation to Atlanta, and championed a promotion to software engineer.
Career stories: Breaking barriers with LinkedIn
Originally from Argentina, systems & infrastructure engineering leader Federico was a founding member of the Media Infrastructure team in 2015.
Now based in Bellevue, Wash., Federico shares how his supportive mentor, LinkedIn’s “sweet spot” scale, and the distinctive engineering challenges here ultimately brought him back to LinkedIn in 2019.
Open sourcing AvroTensorDataset: A performant TensorFlow dataset for processing Avro data
How did LinkedIn Engineering improve data-processing speeds by up to 162x, decreasing overall Machine Learning (ML) training time by up to 66%?
In our latest technical blog post, our deep-learning Systems & Infrastructure team walks us through its solution for loading and processing Avro files via its open-sourced AvroTensorDataset.
This high-performance TensorFlow dataset is designed to be efficient and scalable — making it ideal for large-scale #ML applications — and is available via GitHub here.
New approaches for detecting AI-generated profile photos
With the rise of AI-generated synthetic media and text-to-image generated media, fake profiles have grown more sophisticated.
To keep our platform and 930M+ members safe, LinkedIn Engineering’s AI and Machine Learning (ML) teams have partnered with academia via our LinkedIn Scholars program to research advanced detection techniques, and share our results.
For more on how our jointly developed models have successfully detected 99.6% of a common type of AI-generated profile photos, check out our technical blog post here.
The New Stack: LinkedIn’s real-time graph database Is LIquid
To handle 2.02M graph queries/sec., “we needed [our LIquid database] to be on the Daytona 500 side of the graph databases," shares LinkedIn Engineering Director Bogdan Arsintescu.
In this article from The New Stack, our Systems & Infrastructure team dives into the 4th generation of our LIquid graph — which represents 251B graph edges and 1B nodes.
For more on how we've reduced latencies to 50 ms, and CPU utilization by >3x via LIquid, check out the full technical blog post here.
Authority Magazine: Women reshaping the cybersecurity industry: LinkedIn’s Julie Madhusoodanan on the five things you need to create a highly successful career in the cybersecurity industry
Growing up in India, Julie Madhusoodanan has always been drawn to logic and creating solutions to complex problems, so a career in tech was a clear fit.
In an interview with Authority Magazine, Julie shares more about what’s exciting in the cybersecurity field right now, how she’s working to change misconceptions about cybersecurity, and tips for ensuring you’re protected from breaches.
Find your ideal engineering job today. Check out these, and other engineering jobs at LinkedIn across our 8+ global engineering hubs at lnkd.in/EngJobs:
Mobile (Android/iOS) Senior Staff Engineer — Marketing Identity
Mountain View, Calif. or U.S. Remote | Hybrid or Remote
Staff Software Engineer — Graph Algorithms Platforms
Bellevue, Wash. or U.S. Remote | Hybrid or Remote
Data Senior Engineer — Data Science
Mountain View, Calif. or U.S. Remote | Hybrid or Remote
Powered by the world’s most trusted and largest professional platform, our vision is to create economic opportunity for every member of the global workforce.
Our engineers collaborate cross-functionally to ethically craft quality consumer, business-to-business (B2B) marketing, B2B sales, and talent/learning products — to help job seekers, businesses, and learners maximize their potential.
Shout out to #LinkedIn #Engineering's Xavier (Xavi), Jelanah, Federico, Jonathan, Pei-Lun, Abin, Arup, Lijuan, Shivansh, Gonzalo Aniano, Smit, Bogdan, and Julie who are featured in this month's e-newsletter. 🎉 Hope you all have a fantastic Shutdown Week 🌴 next week!