My research is on developing nonparametric, contextual reasoning models for program synthesis and question answering on knowledge graphs and text. These models are accurate, controllable (debuggable), offer interpretable predictions and can seamlessly reason with newly arriving information. Recently, I have been working on neuro-symbolic advancements to an old nonparametric framework initially proposed in classical AI - Case-based Reasoning. In a CBR framework, the reasoning pattern required to solve a problem are derived from the reasoning patterns of other similar problems. A CBR framework provides a natural way of extending K-nearest neighbor approaches for classification to more complex problems such as program synthesis and question answering. My research interests also include developing models for open-domain QA, building procedural knowledge graphs from text and common-sense reasoning.