The mission of RobustLinks is to use state of the art AI, NLP, and ML to compose what we call Knowledge as a Service (KaaS). The service builds and refines a probabilistic model of user demographic and interests. Given only one or more geolocations KaaS can infer an eleven dimensional probabilistic model of the user demographics. KaaS can also consume unstructured, "snack sized", idiosyncratic textual data (e.g. social media feed) and infer its semantics. It integrates end-to-end technology stack that continuously gather, summarize and represent the semantics of unstructured text, in real-time and at scale. The derived semantics is then analyzed to condense the big data to a compact and searchable semantic graph. We use IR, not semantic web, paradigms and tools. BigSieve is one consumer facing application that we have built on top of KaaS layer. In BigSieve users opt-in to share their location and access to their social media feed. BigSieve then uses KaaS to continuously update a model of the user demographics and interests and match them to items in the application domain, messages or other users.