Senior Python Software Architect - Cogniac

San Jose, CA

Cogniac is a Silicon Valley based enterprise software company with a mission to improve the world by making it easy to automate visual observation tasks.

Cogniac has developed an elegant end-to-end architecture to enable the commercial
deployment of deep convolutional neural networks in enterprise video and image analysis applications. Cogniac System uniquely combines the latest artificial intelligence research with human-computer interaction tools and large-scale data management in order to rapidly and automatically learn to identify what matters most to our business customers so they can take informed, timely and even automated action. Our easy drag-and-drop interface can be used to create any observation workflows very quickly, with very little technical knowledge.

Candidates must have a strong experience with Python. Experience with C/C++ is a strong plus. You must also have a passion for creating simple and elegant microservices. Previous experience in a start-up environment is a huge plus.

Responsibilities:

• Continually conceptualize, design, implement, deploy and improve Cogniac’s innovative AI platform that sustains and enhances customer’s operations.
• Work on the full development life cycle of time sensitive projects.
• Design and build back-end systems and services in Python for our deep learning platform
• Define architecture and best practices
• Write world-class production code and assist in operating it at scale
• Work across teams to architecture new features and platforms.

Requirements:

• 4+ years experience developing Python production code
• Experience with Kubernetes or similar container orchestrators
• Experience with modern cloud-scale development on e.g. AWS, including developing cloud-based APIs and high performance, secure, distributed, multi-tenant services
• Creative attitude and hacker/builder spirit
• Strong communication skills to effectively collaborate with all stakeholders.
• Ability to work efficiently and independently.