In the absence of an efficacious and affordable vaccine, the current crisis of COVID-19 is likely to be a long drawn one for many developing countries. In Bangladesh, where the entire population is susceptible and strict lockdown has been relaxed (as of May 31st 2020) due to concerns over saving livelihoods, the best available resources and capacities in the country have to be mobilized for an integrated and adaptive response strategy. In this paper we argue that a suitable response strategy for a country with highly constrained health system, must consider how response components will be delivered at scale, along with what can be delivered. In order to save maximum number of lives, an optimal strategy will be one that is able to iteratively select the most feasible set of health response and the network of organizations that can deliver most effectively at scale. This might require thinking outside of the conventional vertical network of public health system. Given its history of high-capacity non-government organizations in Bangladesh, it is likely that there are multiple alternative horizontal network options for delivering any set of response interventions. In fact many horizontal networks are already actively engaged in COVID-19 response work. The goal should be to identify and coordinate these networks, create new networks, and embed mechanisms for scaling up what works and scaling down what does not work. For a rapidly escalating and unpredictable crisis such as COVID-19, an adaptive response strategy is needed which allows for old and new networks of organizations to align and work collectively with minimum loss of lives.