Carson Eisenach

Large language models (LLMs) are good at generating code, but remain brittle for formal verification in systems like LEAN4. A core scalability challenge is that verified synthesis requires consistent outputs across multiple artifacts: executable code, precise specifications, theorem statements, and ultimately proofs. Existing approaches rarely treat these as a unified pipeline.

Multi-horizon probabilistic time series forecasting has wide applicability to real-world tasks such as demand forecasting. Recent work in neural time-series forecasting mainly focus on the use of Seq2Seq architectures [25]. For example, MQTransformer [10] – an improvement of MQCNN [27] – has shown the state-of-the-art performance in probabilistic demand forecasting.

In this work, we investigate data-driven solutions aimed towards mitigating the supply risk and optimizing cross-product constraints/resources for the inventory management problem. We particularly focus on dual sourcing methodologies by lead time for inventory planning and control. We investigate how to efficiently employ Deep Reinforcement Learning (RL) models for learning and forecasting the real-world supply chain processes under a range of assumptions.