TRaaS（Technological Risk-defense as a Service）是绿色，稳定，智能的技术风险防控解决方案。TRaaS关注整个研发运维过程可能产生的稳定性风险，从流程制度、文化宣导、技术方案、平台体系多个方面提供稳定性风险防控方案，实现风险的主动发现和自我恢复能力，助力业务高质量增长。
TRaaS 是整个分布式架构和技术风险能力组合在一起的免疫系统，将高可用和资金安全能力结合AIOps，使系统实现故障自愈。此外，TRaaS还具备以下六大特性：
- 统一变更管控，智能变更风险防御；
- 基于chatops的标准SOP故障管理，精细化应急定位辅助；
- 智能资源容量调度，实现稳定性和成本最优平衡；
- 万亿级资金证账实智能实时核对；
- 大规模混沌工程驱动稳定性技术演进，技术风险文化宣导；
- AIOps在可控风险下提升运维效率；
TRaaS (Technological Risk-defense as a Service) is a comprehensive solution for modern Site Reliability Engineering, which prevents and solves all kinds of potential faults and stability risks that arise during the whole software lifetime.
Our main interests lie in the following areas:
- The observability platform that provides various runtime detail of IT infrastructures and businesses.
- Data-driven and intelligent change control system which prevents potential faults.
- Auto-scaling and related facilities, which achieve the optimal balance between stability and resource consumption.
- AIOps infrastructure and its best practice, which aims to improve the efficiency of SREs.
- Large-scale chaos engineering and its best practice.
- CeresDB CeresDB is a cloud-native time-series database that aims to handle both time-series and analytic workloads efficiently.
- HoloInsight HoloInsight is a cloud-native observability platform with a special focus on real-time log analysis and AI integration.
- Pyraformer Low-complexity Pyramidal Attention for Long-range Time Series Modeling and Forecasting This is the Pytorch implementation of Pyraformer (Pyramidal Attention based Transformer) in the ICLR paper
- ICLR 2022 Pyraformer: Low-Complexity Pyramidal Attention for Long-Range Time Series Modeling and Forecasting
- KDD 2022 A Meta Reinforcement Learning Approach for Predictive Autoscaling in the Cloud
- IJCAI 2022 Memory Augmented State Space Model for Time Series Forecasting
- ACM SoCC 2022《DeepScaling: Microservices AutoScaling for Stable CPU Utilization in Large Scale Cloud Systems》
- ICDM 2022 End-to-End Modeling of Hierarchical Time Series Using Autoregressive Transformer and Conditional Normalizing Flow-based Reconciliation
- ESEC/FSE 2022 《Investigating and Improving Log Parsing in Practice》
- ACM International Conference on Management of Data（SIGMOD）2023 《BALANCE: Bayesian Linear Attribution for Root Casuse Localization》
- AAAI 2023 Structured Learning and Task-based Optimization for Time Series Forecasting on Hierarchies
- IJCAI2023《Full Scaling Automation for Sustainable Development of Green Data Centers》