Python Deep Learning Performance/Deployment

Project Description

We want to productionalize an analysis that we've been running offline. For our testing, we built a flask wrapper around an open source package graciously provided by an academic, which unfortunately was not built to scale well. We need someone that has experience with python performance tuning (ideally some familiarity with Theano and flask), and also AWS deployment experience to help us move this to production.

Deliverable: successful deployment of our microservice on AWS.

Costs: for testing, we expect you'll be creating and destroying some GPU instances - these can get expensive. We can add you to our AWS account or reimburse you for the cost of machines on your account (ideal, so you can adjust security settings more freely).

Company: We are a seed-stage startup working on text analytics.