Metronome is seeking a Programmer Analyst to work with a small team, developing a framework and tools to enable cutting edge artificial intelligence experiments for NGA’s Research Directorate. Activities include:
Using the Jextr application to:
* Develop ARF’s tagging and labeling tool
* Gather user feedback as v1.0 is deployed to a larger base, and identify any bugs found and begin addressing those ahead of a v2.0 release
* Continue working through the existing backlog of gitlab issues which have been slated for a future release (include features like added security for private projects, better auto-segmentation, … )
* As new projects are identified for experiments with ARF, ingest any existing labeled training data they may be using into JEXTR database
Perform ARF Experiments:
* Review, edit, and further develop existing code provided by ARF’s experiment partners
* Explore different AI and MLOps platforms to be used within ARF
* Document all findings on ARF Confluence page
* Develop model training and inferencing templates which can be used for demonstrations of a platforms strengths and capabilities, or for a new project to use as boiler plate start-up code
Manage API Endpoints to:
* Identify the API endpoints for the software and platforms that ARF uses
* Develop scripts, set AWS roles, and validate permissions to allow for automated data flows between those API endpoints (e.g., Being able to train a model in Syntasa, and save it to the Deepcore core model catalog)
* Python coding (intermediate skills are fine, as the majority of python work is reviewing existing code and editing on top of it. Not too much requirement to build from scratch. 2-3 years minimum of python specific should be fine).
* Familiarity with Machine Learning model training and inferencing. Should understand these elements: Training Datasets (curating them, tracking them, storing them), Model training in both pytorch and tensorflow, Hosting model inferencing through cloud processing
* AWS Cloud Engineering / Architect skills are going to be important moving forward. I’d look for someone with the AWS Cloud Practitioner cert as an easy flag, although having the certifications isn’t a requirement if they have the experience / knowledge
* Linux experience, or shell script experience
* Machine Learning Operations experience
* Database Engineering