about me

Like the title says, my name's Robert Brown.  I'm a Program Manager at Microsoft, working on speech technologies - in particular, the APIs.

I expect this log to be about both program management and speech.  We'll see if I predict correctly.

What's speech?  It boils down to two main functions: synthesis (an app provides text, which is converted into a synthetic human voice); and recognition (a person speaks, and the app receives a transcript of what they said).  Of course, that's like saying computer graphics boils down to a grid of pixels where you just paint each one the right colour and voila you have Halo 2.  The point is that I started working in speech in 2002, which isn't so long ago, and while a few years is enough to develop a healthy knowledge of the topic, it's one of those fascinating areas that just gets more fascinating as I learn more about it.  So I thought a blog might be a good way to share some of these thoughts.  It may also a nice way to chat about new developments when they're announced.

What's program management?  IMO probably the least clearly defined job there is, and I'd say that's by design.  Developers write the code in a product and fix any bugs.  Testers test that the product does what the spec says it should do and log any bugs.  Both are demanding jobs, but at least they're easy to describe. 🙂  Program managers do a bunch of stuff that I generally sum up as "making sure we ship the right stuff at the right time", which involves varying degrees of: process; planning; writing functional specs; researching customers and the industry; acting as a "randomization" buffer so the devs and testers can get on with producing the right lines of code with minimal distraction; representing the team to other teams, upper management; etc.  I've had a few different program management jobs, and the mix has been different in each one.  I also didn't see any topics like this when I was at University (class of '91).  So it's all been "professional development" - i.e. learn by doing.  If this were karate, I guess I'd be one of those belts in the middle (red? - I don't actually know the colours), which means I can take care things just fine, but that black belt's still waiting for me.  I figured it'd be interesting to post anything I learn on the way.

Comments (2)

  1. Edward says:

    I hope Longhorn has a better synth than Microsoft Sam, it seems so pointless to have all those easy to use APIs when the end result is unlistenable.

    I know you can plug into other SAPI compliant synths, but they arn’t that easy for end users to get hold of, and it would be good to have at least one acceptable voice out of the box.

  2. kamaleswaran says:

    If u r a red belt then i’m no way near to white belt(begineers) just going to start learning about this speech server technology.

Skip to main content