Scalability remains a critical challenge in blockchain-based electronic voting systems (BBEV), in particular when it comes to organizing large-scale elections with millions of voters. This paper presents a BBEV integrating state-channels, commit-chains and Directed Acyclic Graphs (DAGs) to address scalability challenges in large-scale elections. The architecture employs regional servers for distributed vote handling and implements aggregated transactions through the Fantom platform’s DAG-based consensus mechanism. Results indicate the system achieves a throughput of up to 10,000 concurrent votes with 10 regional servers and scales as the number of servers increases. Latency remains consistently low, with end-to-end average vote processing time under 2300 milliseconds (±1612 ms) even at maximum concurrency levels. Cost efficiency is significantly enhanced through the aggregation process, reducing the average cost per vote by 99% with a batch size of 100 votes compared to individual vote submissions. Additionally, the system maintains absolute efficiency, as every vote cast is recorded on the blockchain.