Beta is out!

OK, so most of you probably know this already but the Microsoft Speech Server Beta was released on Monday.  This Beta includes several tools that until now I was not allowed to blog about publicly.

The two most interesting tools are a Windows Workflow based dialog flow designer and a conversational grammar builder.  The conversational grammar builder (CGB) allows you to author single slot and multi slot grammars in a much easier way than the existing grammar editor (which still exists) as well as allows you to develop HMIHY grammars ("How May I Help You").

I must admit at first I was a bit skeptical of authoring speech applications through workflow.  I preferred the code approach - writing apps in the Core API where everything is under my control.  There are still times when you must go to the Core API, and I will discuss some of these in later posts.  But after developing some apps in the DFD (Dialog Flow Designer) and seeing apps others have written, I am convinced this is the best way to author apps.  There are still some peculiarties with the designer, but I hope we will have these ironed out before we ship.

I really think MSS will change speech application development significantly.  No more learning complicated markup languages such as VXML or SALT.  If you know how to write code in .NET, you can develop any type of telephony based speech application.

Comments

• Anonymous
  May 12, 2006
  I agree that this will really change the way we write apps.
  
  Though there is  a part of me that thinks this could be bad - now everyone will be writing speech apps and I'm afraid that they will produce some really awful stuff. There is more to a good speech app than just code.
  
  See you next week.
• Anonymous
  May 15, 2006
  What type of hardware is required for Speech Server 2007?
• Anonymous
  May 15, 2006
  Thanks for the update Joe.  We were accepted into the Beta program, but I haven't received any info on when/how we can download the beta.  Do you know when we can expect this?