This study analyses the relational dimension and the knowledge transfer mechanisms in an innovation ecosystems (IEs), assuming that the bottom-up creation of synergies and cooperative mechanisms between local actors are the drivers of a regional smart growth. More specifically, the study explores the configuration of the network structure and the variety of inter-organizational relationships in a case of a smart IE by capturing the heterogeneous nature of IE demography, whether most studies limit their analyses to inter-firm relationships and at the node-level. Secondly, the paper provides insights into the network portfolio composition, which has been underexplored in IE literature, allowing for the identification of those relationships considered more fruitful to enhance innovation processes from a local perspective. To capture both aspects of IE’s relational dimension (i.e. network structure and network portfolio of relationships) our paper adopts an explorative approach, by taking evidence from the empirical study of the biopharma IE in greater Boston area, which has been exemplified as a successful case. Our empirical study combines two methods, namely social network analysis and expert interviews. Firstly, we conduct a social network analysis to gain insights about the optimal network structure and secondly, we conduct a round of semi-structured interviews with key stakeholders in the ecosystem to explore the characteristics of the desirable network portfolio. Our findings show that a smart IE presents an open network structure with structural holes, a high level of modularity and a portfolio of relationships that privileges informal and non-redundant ties within small communities focused on specific themes.