Aquaculture is experiencing remarkable growth, particularly in South East Asia. In this region, the rearing of minimum two species at the same time (polyculture) in ponds is largely widespread. It improves the resource use, fish welfare and the system resilience. Nonetheless, this system is complex not least because wild fish population (associated biodiversity) can enter into the pond. Therefore, potential benefits can be expected provided the fish introduced by the farmers (planned biodiversity) and the associated biodiversity are compatible (i.e. able to live in the same system without detrimental interaction and with a minimized competition) or even complementary (i.e. able to use different portions of available resources or display commensal/mutualistic interactions). Therefore, it is important to identify the most promising combinations of species that may be introduced for a better productivity in this open system. Compatibility is a prerequisite for any polyculture in an aquaculture system. However, its evaluation cannot be carried out by empirical testing alone for pragmatic reasons (e.g. considering 7 species, 120 possible combinations of 2, 3, 4, 5, 6 and 7 of these species would have to be tested) and ethical reasons (i.e., welfare concern by combining not compatible species). One solution lies in a preliminary in silico assessment of compatibility without bioassays, so as to focus experimental evaluations or field tests solely on the most promising combinations. Informatic tool (i.e., AquaDesign) has been already developed to assess the abiotic compatibility (i.e., ability of living in the same abiotic environment) of given fish combinations. However, there is currently no tool available to evaluate the biotic compatibility between species (i.e. ability of living together system without detrimental interaction and with a minimized competition). We thus introduce an in-silico model for assessing biotic compatibility using fish functional traits. We apply it for the improvement of fish polyculture in pond systems in Cambodia. The rearing system will consist of cages with Nile tilapia (Oreochromis niloticus) within a pond containing associated biodiversity that comes from flood during the rainy season and 2 or 3 species from the planned biodiversity. The implemented data in the model are functional traits from scientific literature mined in the database TOFF. The outcome of the model is a compatibility index for each combination that incorporate the risk of predation and competition for trophic and spatio-temporal resources. In the context of Cambodia aquaculture, the next step will be to select combinations based on socio-economic expectations before performing field tests to validate the model’s predictions against real-world outcomes.