Cortex
Cortex, a prominent reseller, system integrator, and consulting company, that plays a vital role in numerous system integration and digital transformation initiatives. Cortex collaboration with software players such as Google empowers organizations to embrace disruptive technologies and accelerate their journey to the cloud, thus reshaping their business models.
| Logo | Customer | Industry | Empl. | Revenue | Country | Vendor | Product | Category | When | Insight |
|---|---|---|---|---|---|---|---|---|---|---|
|
|
Twitter, Inc. | Media | 7500 | $5.1B | United States | Google TensorFlow | ML and Data Science Platforms | 2017 |
In 2017, Twitter, Inc. adopted Google TensorFlow within its ML and Data Science Platforms work to retool ranking for the home timeline and centralize model development. The decision was made in late 2017, with Twitter describing a migration that positioned TensorFlow as the primary training and serving stack for timeline relevance models, and Cortex engaged as a collaborator on the effort.
The implementation focused on expressing the entire model graph for home timeline relevance as programmatic Python functions, enabling development of complex relevance models that score thousands of candidate Tweets per user. The deployment encompassed both training pipelines and a serving stack for production scoring, with models designed to predict engagement and signals that encourage healthy public conversation. Google TensorFlow supported both immediate production use cases and ongoing research experiments, reflecting the ML and Data Science Platforms category capability set for model training, graph definition, and serving.
Operational integration emphasized interoperability with Twitter production systems that are predominantly JVM based, leveraging TensorFlow Java APIs to connect model serving into existing infrastructure. The migration replaced Twitter’s prior Lua Torch based platform with a TensorFlow training and serving architecture, after evaluating alternatives such as PyTorch. Cortex and Twitter’s central machine learning and AI team worked together on the stack design and rollout, aligning training workflows, model artifact formats, and production scoring endpoints.
Governance and process changes centered on standardizing model development in Python, accelerating experiment iteration, and instituting a TensorFlow based workflow for research to production handoff. Twitter cited improved developer productivity and stronger access to industry research as explicit outcomes from adopting Google TensorFlow, while retaining production stability guarantees that matched the scale requirements of serving hundreds of millions of active users.
|
| First Name | Last Name | Title | Function | Department | Phone | |
|---|---|---|---|---|---|---|
| No data found | ||||||
Buyer Intent: Companies Evaluating Cortex Services
| Logo | Company | Industry | Employees | Revenue | Country | Evaluated | ||
|---|---|---|---|---|---|---|---|---|
| No data found | ||||||||