AI • September 20, 2023

Quantiphi Joins the TensorFlow AI Service Partner Program to Lead Enterprise Adoption of AI

Artificial intelligence (AI) and machine learning (ML) technologies are transforming the way we do business, from delivering operational efficiency to developing a better understanding of customers and markets. Implementing AI at scale provides businesses with strategic advantage over the competition by reducing risks, optimizing costs, and increasing revenue. 

Businesses and enterprises, however, often require the expertise from external resources to successfully implement and take advantage of ML solutions within their workflows. We are excited to announce that TensorFlow has selected Quantiphi as a Trusted Partner for ML-led business transformations. This partnership is a recognition of our expertise in developing and deploying ML solutions. Quantiphi’s in-house TensorFlow & Google Cloud Platform certified professionals translate state-of-the-art research into production-ready solutions, allowing businesses to harness the power of latest advances in AI and ML. 

As one of the early adopters of TensorFlow, Quantiphi has integrated the ML models into customers’ end-to-end workflows to transform and scale operations. The association that began in 2016 led to the implementation of distributed training on Cloud AI platforms with up to 32 GPUs, using a hybrid distribution strategy that resulted in higher speed. We also trained the ASR models on up to 10,000 hours of data.

Quantiphi has been using Tensorflow as a platform for building enterprise ML solutions for wide-ranging applications like medical imaging, video analytics, and natural language understanding. We have leveraged the flexibility and adaptability of TensorFlow workflows to integrate ML models in innovative applications across technologies. We incorporated GPU based RNN cells and reduced strategy for gradient aggregation with speed on GPUs, which improved performance by reducing the training time. 

Quantiphi recently used TF for training a DistillBERT model on custom datasets. The trained model powered a Question and Answering system for mining insights from pharmaceutical documents. By using Tensorflow Serving for deployment, Quantiphi attained a sub-second inference time that enhanced the user experience and made the project a huge success. 

One of the innovative solutions included the use of TF.js to embed a Pose Estimation and Classification model in the browser for real-time detection of gestures and emotions directly from a user’s computer camera feed. Another application was built for writing custom VNet-style architectures for the development of 3D image segmentation systems that can capture lung CT scans and accurately identify abnormalities. TF Estimators and TPUEstimators (and other TPU supported ops) utilized to train models on multiple GPUs and TPUs respectively helped Quantiphi to scale the model training process and expand the benefits of the technology for its customers.

Quantiphi leveraged TF Serving to deploy the models on Kubernetes clusters and improve the efficiency of the deployment process. The TensorFlow ecosystem offers tools that support the development of applications from the early stages of applied research to their deployment and monitoring. TensorFlow provides the framework to build enterprise ML solutions that enable continuous evolution of the models with the data distribution. 

More Resources:

Learn more about our other partnerships with Google Cloud Platform and Amazon Web Services.
Check out some of our customer use cases and success stories
Ready to talk about transforming your business? Connect with us today.

Written by


Thank you for reaching out to us!

Our experts will be in touch with you shortly.

In the meantime, explore our insightful blogs and case studies.

Something went wrong!

Please try it again.