Ramachandra Vikas Chamarthi

• Productionalized Proscia's First Research focused AutoQC product used OpenVINO based acceleration acceleration for optimized CPU based pipeline improving latency by 30%.
• Worked on transforming old AWS cloudformation AutoQC infrastructure into Kubernetes infrastructure making system more reliable and available.
• Enabled Research Pipeline to support Philips iSyntax WSI format (Only FDA supported pathology WSI format).
• Worked closely with two partner lab sites to perform extensive beta study to showcase the Utility of DermAI Clinical Triage product.
• Major tech debt reduction by complete overhaul existing code base Python3.8 and updated internal packaging tool mechanism to using Git CI, poetry, AWS CodeArtifact.

• Part of the founding Al team got the opportunity to work on Al POC applications and research pipeline acceleration using internal validation systems. Perform beta deployments aiding utility studies.
• Worked on R&D POC AUTOQC product a pathology slide quality product from development to productionalizing end to end first prototype with AWS-CloudFormation, RabbitMQ, Docker.
• Accelerated pipeline components with efficient in-house C++ python modules and other industry optimization libraries like TensorRT to generate efficient pipeline iterations. Accelerated Proscia first AI product DermAI by 40% and also enabled around 80% of cost saving in Data preparation processes replacing compute intensive GPU module into optimized CPU executable iterations enabling spot instance usage reducing AITeam AWS Ec2 costs from 20,000 USD / Month to 400 USD / Month.

• Worked on chemical property prediction deep learning algorithms to showcase the usability of ML-Suite of WOLFRAM MATHEMATICA.

• Worked on establishing aws/onprem hybrid compute infrastructure to aid research efforts. Analyzed ec2 compute costs to replace median GPU compute workload with a onprem machine until 2022 with a cost 80K USD Machine effectively offloaded ec2 compute costs of 200K USD and will continue to do so until the life of machine.

Member of TECSAR UNCC, worked on various projects based on GPU Architecture | GPU Simulators | Deep Learning | Cache replacement policies e.t.c.

MS In EE from Dept of Electrical and Computer Engineer.

With Thesis: Algorithmic optimization of the first convolution layer in CNN for hardware accelerator design.