TimeCraft offers a unified, practical solution for real-world time series generation—combining cross-domain generalization, text-based control, and task-aware adaptation. It’s designed to produce high-quality, controllable synthetic data that’s both realistic and useful for downstream applications. 

More details can be found here: https://github.com/microsoft/time-craft 

PIKE-RAG enhances Retrieval Augmented Generation (RAG) systems by focusing on extracting, understanding, and applying domain-specific knowledge while building coherent reasoning logic. It addresses challenges in knowledge segmentation and retrieval accuracy by using context-aware segmentation techniques and automatic term label alignment. PIKE-RAG excels in complex reasoning tasks, demonstrated by its high performance on multi-hop question answering datasets like HotpotQA and MuSiQue. It has shown significant improvements in fields such as industrial manufacturing, mining, and pharmaceuticals, making it a versatile tool for integrating multi-source information and performing multi-step reasoning. 

More details can be found here: https://github.com/microsoft/pike-rag 

RDAgent aims to automate the most critical and valuable aspects of the industrial R&D process, and we begin with focusing on the data-driven scenarios to streamline the development of models and data. Methodologically, we have identified a framework with two key components: 'R' for proposing new ideas and 'D' for implementing them. We believe that the automatic evolution of R&D will lead to solutions of significant industrial value.

R&D is a very general scenario. The advent of RDAgent can be your

• 💰 Automatic Quant Factory (🎥Demo Video|▶️YouTube)

• 🤖 Data Mining Agent: Iteratively proposing data & models (🎥Demo Video 1|▶️YouTube) (🎥Demo Video 2|▶️YouTube) and implementing them by gaining knowledge from data.

• 🦾 Research Copilot: Auto read research papers (🎥Demo Video|▶️YouTube) / financial reports (🎥Demo Video|▶️YouTube) and implement model structures or building datasets.

• ...

You can click the links above to view the demo. We're continuously adding more methods and scenarios to the project to enhance your R&D processes and boost productivity.

Additionally, you can take a closer look at the examples in our 🖥️ Live Demo.

More details can be found here: https://github.com/microsoft/rd-agent 

We introduce MineWorld, an interactive world model on Minecraft that brings several key advancements over existing approaches:

• 🕹️ High generation quality. Built on a visual-action autoregressive Transformer, MineWorld generates coherent, high-fidelity frames conditioned on both visuals and actions.

• 🕹️ Strong controllability. We propose benchmarks for the action-following capacity, where MineWorld shows precise and consistent behavior.

• 🕹️ Fast inference speed. With Diagonal Decoding, MineWorld achieves a generation rate of 4 to 7 frames per second, enabling real-time interaction in open-ended game environments.

More details can be found here: https://github.com/microsoft/mineworld

MarS is a cutting-edge financial market simulation engine powered by the Large Market Model (LMM), a generative foundation model. MarS addresses the need for realistic, interactive, and controllable order generation. This paper's primary goals are to evaluate the LMM's scaling law in financial markets, assess MarS's realism, balance controlled generation with market impact, and demonstrate MarS's potential applications. 

More details can be found here: https://github.com/microsoft/mars 

We introduce VidTok, a cutting-edge family of video tokenizers that excels in both continuous and discrete tokenizations. VidTok incorporates several key advancements over existing approaches: ⚡️ Efficient Architecture. Separate spatial and temporal sampling reduces computational complexity without sacrificing quality. 🔥 Advanced Quantization. Finite Scalar Quantization (FSQ) addresses training instability and codebook collapse in discrete tokenization. 💥 Enhanced Training. A two-stage strategy—pre-training on low-res videos and fine-tuning on high-res—boosts efficiency. Reduced frame rates improve motion dynamics representation. VidTok, trained on a large-scale video dataset, outperforms previous models across all metrics, including PSNR, SSIM, LPIPS, and FVD. Blog: https://www.microsoft.com/en-us/research/blog/vidtok-introduces-compact-efficient-tokenization-to-enhance-ai-video-processing/ 

More details can be found here: https://github.com/microsoft/VidTok

The performance degradation of lithium batteries is a complex electrochemical process, involving factors such as the growth of solid electrolyte interface, lithium precipitation, loss of active materials, etc. Furthermore, this inevitable performance degradation can have a significant impact on critical commercial scenarios, such as causing 'range anxiety' for electric vehicle users and affecting the power stability of energy storage systems. Therefore, effectively analyzing and predicting the performance degradation of lithium batteries to provide guidance for early prevention and intervention has become a crucial research topic.

To this end, we open source the BatteryML tool to facilitate the research and development of machine learning on battery degradation. We hope BatteryML can empower both battery researchers and data scientists to gain deeper insights from battery degradation data and build more powerful models for accurate predictions and early interventions.

Our paper is now available on Arxiv and ICLR 2024! This paper provides detailed introduction to our design, which we will be actively updating during the development of BatteryML. 

More details can be found at the site https://github.com/microsoft/batteryml.

A wide range of industrial applications desire precise point and distributional forecasting for diverse prediction horizons. ProbTS serves as a benchmarking tool to aid in understanding how advanced time-series models fulfill these essential forecasting needs. It also sheds light on their advantages and disadvantages in addressing different challenges and unveil the possibilities for future research.

To achieve these objectives, ProbTS provides a unified pipeline that implements cutting-edge models from different research threads, including:

• Long-term point forecasting approaches, such as PatchTST, iTransformer, etc.

• Short-term probabilistic forecasting methods, such as TimeGrad, CSDI, etc.

• Recent time-series foundation models for universal forecasting, such as TimesFM, MOIRAI, etc.

Specifically, ProbTS emphasizes the differences in their primary methodological designs, including:

• Supporting point or distributional forecasts

• Using autoregressive or non-autoregressive decoding schemes for multi-step outputs

More details can be found at the site https://github.com/microsoft/probts.

Qlib is an AI-oriented quantitative investment platform, which aims to realize the potential, empower the research, and create the value of AI technologies in quantitative investment.

It contains the full ML pipeline of data processing, model training, back-testing; and covers the entire chain of quantitative investment: alpha seeking, risk modeling, portfolio optimization, and order execution.

With Qlib, user can easily try ideas to create better Quant investment strategies.

For more details, please refer to the site https://github.com/microsoft/qlib and our paper "Qlib: An AI-oriented Quantitative Investment Platform".

Multi-Agent Resource Optimization (MARO) platform is an instance of Reinforcement learning as a Service (RaaS) for real-world resource optimization. It can be applied to many important industrial domains, such as container inventory management in logistics, bike repositioning in transportation, virtual machine provisioning in data centers, and asset management in finance. Besides Reinforcement Learning (RL), it also supports other planning/decision mechanisms, such as Operations Research. 

More details can be found at the site https://github.com/microsoft/maro. 

FOST (Forecasting open-source tool) aims to provide an easy-use tool for spatial-temporal forecasting. The users only need to organize their data into a certain format and then get the prediction results with one command. FOST automatically handles the missing and abnormal values and captures both spatial and temporal correlations efficiently. 

More details can be found at the site https://github.com/microsoft/FOST.