The AI4Code at IBM and Red Hat talk aims to showcase the latest in AI being applied to the code and programming lifecycle as relevant to enterprise applications. The session will feature live demos and perspective of research from IBM and Red Hat that currently use AI and ML techniques to make the entire code lifecycle more efficient, scalable, creative, and secure from an industry research perspective.
CodeBreaker is a coding assistant for data science code. It performs domain specific knowledge extraction over corpora such as GitHub, data science ontologies and Wikipedia, and documentation about data science code and tutorials to create a knowledge graph (KG) related to data science code. This KG is then used in downstream products like IDEs, etc. in order to make writing ML code easier.
Red Hat’s Project Thoth focuses on analyzing and recommending software stacks for AI applications. The team will demonstrate how Thoth learns new knowledge and uses that knowledge with reinforcement learning techniques to recommend application stacks.
Code and app modernization is a core problem for the software industry, with a large amount of legacy code running critical systems world over. The "Intelligent Application Insights" tool builds and customizes models that generate containerization strategies for a given application; and constructs customized cost models for cost/benefit and risk assessments based on dynamic Bayesian learning.
AI for Vulnerability Analysis (AI4VA) models source code as a graph neural network in order to map that code to potential software vulnerabilities. Specifically, it learns whether signatures of the vulnerabilities in source code can be learned from their graph representation.
Kartik Talamadupula (IBM Research)
Julian T Dolby (IBM Thomas J. Watson Research Center)
Kavitha Srinivas (IBM Research)
Researcher with expertise in semantic technologies, knowledge graphs, linked data, data integration.
Fridolín Pokorný (Red Hat)
Maja Vukovic (IBM Research)
Anup K Kalia (IBM T. J. Watson Research Center)
Alessandro Morari (IBM)
More from the Same Authors
2021 Datasets and Benchmarks: Dataset and Benchmark Track 1 »
Joaquin Vanschoren · Serena Yeung · Maria Xenochristou
2020 Workshop: KR2ML - Knowledge Representation and Reasoning Meets Machine Learning »
Veronika Thost · Kartik Talamadupula · Vivek Srikumar · Chenwei Zhang · Josh Tenenbaum
2019 Workshop: KR2ML - Knowledge Representation and Reasoning Meets Machine Learning »
Veronika Thost · Christian Muise · Kartik Talamadupula · Sameer Singh · Christopher Ré