The efficiency of decentralized book systems like Bitcoin and Ethereum has always been a challenge. It is usually measured by three major factors: scalability, throughput, and latency. Scalability refers to how the system capacity is increased by adding more physical resources. Throughput measures the volume of transactions for a given period of time, where most current solutions attempt to improve such as NEO, EOS, etc. Latency measures the processing time of any single transaction. In current blockchain based systems, the block generation rate is the main latency bottleneck. Off-chain processes such as state channels are the most recent work that can integrate partial inbound transactions, reducing latency. Unfortunately, the state channel introduces more issues at the same time, such as cross-channel synchronization, which makes the state channel unavailable for full adoption of current blockchain solutions. In order to solve the efficiency problem, we proposed an end-to-end solution called ALZA, which links the dedicated high-throughput blockchain with self-organizing payment fields. This mechanism allows arbitrary set of users to create payment fields that process extremely low latency transactions within each field. Therefore, users can make transactions almost immediately. Since all transactions are conducted within fields, transaction costs will be reduced by several orders of magnitude. In addition, ALZA distributes main ledger to each client through an innovative replication mechanism. Therefore, the system will be significantly more robust to blockchain system failures. In theory, ALZA can complete millions of transactions in one second, which naturally supports high-frequency trading.