(This doesn't really answer your question, but hopefully is still useful)
I actually do this for startups, but it's a bit of a specialized and difficult art. Answering specific questions about technology stacks requires having used them before, and most people only have experience with a small subset of them (myself included, so I probably couldn't answer half of your questions). As a result, I often break down the question into phases.
First, broader questions on design are actually a lot easier to answer. For example: "for problem X, should I use a producer/consumer model + sharded key/value or a direct RPC to a search stack?". Or even more open ended: "on this budget, how can I consume the twitter firehose in realtime to do X?".
Second, the more specific "should I use RabbitMQ or Kafka" for my producer/consumer component requires specialized knowledge. Once you break your question down to that level of granularity, though, it is relatively easy to find an expert.
I actually do this for startups, but it's a bit of a specialized and difficult art. Answering specific questions about technology stacks requires having used them before, and most people only have experience with a small subset of them (myself included, so I probably couldn't answer half of your questions). As a result, I often break down the question into phases.
First, broader questions on design are actually a lot easier to answer. For example: "for problem X, should I use a producer/consumer model + sharded key/value or a direct RPC to a search stack?". Or even more open ended: "on this budget, how can I consume the twitter firehose in realtime to do X?".
Second, the more specific "should I use RabbitMQ or Kafka" for my producer/consumer component requires specialized knowledge. Once you break your question down to that level of granularity, though, it is relatively easy to find an expert.