Along with the thriving of digital technology, the rise of the gig economy has spawned a great number of platform workers. The online labor platforms use algorithm to manage platform workers, which make the platform workers work in a human-machine interaction environment without human connections. Previous studies have mainly focused on the passive experiences of platform workers under the algorithm control and paid less attention to platform workers’ possible proactive adaption. However, platform workers do not only passively accept the algorithm control, but also have the initiative to adapt to the algorithmic working environment. Additionally, existing research has mainly focused on the functional nature of algorithm control, investigating how algorithm control influences platform efficiency and employees’ experiences. The new work environment of human-machine interaction is lack of social connections, which has changed platform workers from “social people” to “system people”. Relational need is one of the basic needs of human beings. Establishing high-quality relationship with others in the workplace can promote employees’ job performance and well-being. It is necessary to investigate the relational nature of the algorithm and investigate platform workers’ behaviors under the human-machine interaction work situation that has almost no human intervention.Based on the proactive motivation model and person-environment fit theory, this study explored the indirect relationships between perceived algorithm control and (a) service performance and (b) burnout through relational crafting and their boundary conditions. According to proactive motivation model, when the working conditions cannot fit the personal needs of employees, individuals will engage in proactive personal-fit behaviors to change their status quo and adapt to the environment requirements. Relational needs are the basic needs of individuals. Work environment of human-computer interaction makes platform workers become independent workers and lack of social connections, which stimulates their proactive behaviors (i.e., relational crafting) to meet relational needs. Further, based on the person-environment fit theory, on the one hand, platform workers can better adapt to the algorithm control environment through relational crafting to satisfy their relational needs, and further improve service performance. On the other hand, the interactions between platform workers and customers can expand their social networks and improve job satisfaction, and further reduce work burnout efficiently. Additionally, self-efficacy can not only moderate the relationship between perceived algorithm control and relational crafting but also the indirect relationships between perceived algorithm control and (a) service performance and (b) burnout through relational crafting. Workers who have high self-efficacy will be more proactive in adapting to the environment and satisfy their relational needs through relational crafting, which can ultimately improve service performance and decrease work burnout.We conducted our study in a logistics platform in western China, used both primary data and secondary data from the platform, and obtained multi-wave and multi-source survey data. We collected 220 sets of data from the delivery workers and we analyzed our hypotheses through MPLUS 7.4.The results show that platform workers perceived algorithm control had a significant positive effect on relational crafting, and relational crafting mediated the relationships between perceived algorithm control and (a) service performance and (b) burnout. Additionally, self-efficacy moderated the indirect relationships between perceived algorithm control and (a) service performance and (b) burnout through relational crafting. Specifically, platform workers with high level of self-efficacy are able to adapt to the work environment through relational crafting behaviors, and further reduce work burnout and increase their service performance.Our research makes some contributions to the literature. First, this study explores the influence of algorithm control among platform workers through the lens of individual proactivity. We investigate how platform workers proactively adapt to the human-machine interaction work context. Second, this study focuses on the relational nature of algorithm control and investigates platform workers relational crafting behaviors when facing human-machine interaction working environment, which also fills the gap of current literature that mainly research about the functional nature of algorithm control. Finally, this research enriches the antecedents of job crafting. Human-machine interaction under the algorithm control is a new working environment. Our research investigates how perceived algorithm control triggers platform workers relational crafting behavior.Additionally, our research provides suggestions for both platform workers and platform companies. As for the platform workers, other than passively accepting the rules of the algorithm, they can also adapt to the environment through proactive behaviors. As for platform companies, they should focus not only on improving the efficiency of work and saving human capital, but also on improving the work experiences of platform workers. Platform companies can provide opportunities for platform workers and encourage them to be initiative to increase service performance as well as their well-being.