MorphingDB: A Task-Centric AI-Native DBMS for Model Management and Inference

2025-11-26

AI Summary

Existing AI-native databases face two key bottlenecks: model-centric paradigms require labor-intensive manual configuration, raising development costs; task-centric AutoML approaches incur high computational overhead and exhibit weak integration with DBMSs. This paper introduces the first PostgreSQL-embedded AI-native DBMS, adopting a task-centric architecture that automates deep learning model storage, selection, and inference for time-series, text, and image tasks. We propose novel contributions: a multidimensional tensor data type; a two-stage transfer learning framework; a pre-embedding sharing mechanism; and a DAG-based batched inference pipeline. Implemented via LibTorch extensions, our system supports BLOB-based model storage, feature-aware mapping, vectorized sharing, and cost-aware scheduling. Evaluated on nine public datasets, it achieves 3.2× higher inference throughput and reduces GPU memory usage by 57% versus state-of-the-art AI-native DBMSs and AutoML platforms—while maintaining competitive accuracy, latency, and resource efficiency.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Data Mining & Knowledge Management: Intelligent Query Processing
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

The increasing demand for deep neural inference within database environments has driven the emergence of AI-native DBMSs. However, existing solutions either rely on model-centric designs requiring developers to manually select, configure, and maintain models, resulting in high development overhead, or adopt task-centric AutoML approaches with high computational costs and poor DBMS integration. We present MorphingDB, a task-centric AI-native DBMS that automates model storage, selection, and inference within PostgreSQL. To enable flexible, I/O-efficient storage of deep learning models, we first introduce specialized schemas and multi-dimensional tensor data types to support BLOB-based all-in-one and decoupled model storage. Then we design a transfer learning framework for model selection in two phases, which builds a transferability subspace via offline embedding of historical tasks and employs online projection through feature-aware mapping for real-time tasks. To further optimize inference throughput, we propose pre-embedding with vectoring sharing to eliminate redundant computations and DAG-based batch pipelines with cost-aware scheduling to minimize the inference time. Implemented as a PostgreSQL extension with LibTorch, MorphingDB outperforms AI-native DBMSs (EvaDB, Madlib, GaussML) and AutoML platforms (AutoGluon, AutoKeras, AutoSklearn) across nine public datasets, encompassing series, NLP, and image tasks. Our evaluation demonstrates a robust balance among accuracy, resource consumption, and time cost in model selection and significant gains in throughput and resource efficiency.

Problem

• Automates model storage selection and inference within PostgreSQL databases
• Reduces development overhead and computational costs in AI-native DBMS
• Optimizes inference throughput while maintaining accuracy and efficiency

Innovation

• Automates model storage and inference in PostgreSQL
• Uses transfer learning for efficient model selection
• Optimizes throughput with pre-embedding and batch pipelines