The effect of climate change, through rising temperature, incidents of drought, and unusually high-intensive rains at times on crop development is well known. However, the conventional methods of mitigating these abiotic stresses in the Indian context largely include post-facto interventions such as proportionate compensation to farmers in the form of weather-based insurance, infrastructure-intensive goals such as improvement of irrigation systems, or promotion of tolerant varieties and resilient practices. In the given context of increasingly changing weather patterns, the unpredictability is quite high making post-facto solutions less promising and sustainable. In this chapter, we propose that machine learning applications can play a bigger role in mitigation and adaptation to abnormal weather patterns supplementing traditional methods. Going forward, we highlight how a policy framework for this can be developed to effectively tackle the emerging challenge. Finally we use case studies of existing practices to show where such a policy framework can fit in.