It's hard to believe, at least for me, that SAPI 5 is 5 years old now. For those of you that don't know, SAPI 5 is the Microsoft Speech API, version 5, that was released to the world on 10/19/2000.

SAPI 5 has been the basis for speech application development inside Microsoft since that time. Office used SAPI 5 to speech enable Office XP in 2001 (3/2/2001). Tablet PC used SAPI 5 to speech enable Windows XP Tablet PC Edition in 2002 (11/10/2002). And we're using SAPI 5 to speech enable all client SKUs of Windows Vista right now!

SAPI 5's release had 4 main components:

  • Microsoft’s version 5 SR engine, code named Whisper
  • Microsoft’s version 5 TTS voices, Mike, Mary, and Sam
  • Microsoft’s version 5 of the platform APIs, typically referred to as SAPI
  • The brand new SAPI 5 SDK, that included samples and tools to help create speech enabled applications.

The engine is currently on version 8 for Windows Vista, but the platform has remained largely unchanged (from a surface area perspective). There was a release of SAPI 5.1 and then a small dot release for Windows XP that both included the ability to call SAPI from automation clients, such as Visual Basic, or J-Script.  The SDK has also had small changes along the way. There were a couple other small releases of SAPI, after SAPI 5.1.

I’m quite fond of SAPI 5. Shipping SAPI 5 on 10/19/2000 was the proudest day for me in my career at Microsoft up to that point. I’ll remember that day for a long, long time.

It was the culmination of a lot of hard work by a lot of people. Fil Alleva, Mark Nickle, Paco Galanes and I were all development leads on the project. Fil was in charge of the core engine, Mark owned the SDK team, Paco owned the TTS team, and I lead the SAPI 5 runtime team. Ralph Lipe, Ed Connell, Philipp Schmid, Rich Pletcher, Dave Wood, Adrian Garside, and Yunus Mohammed were all on the SAPI 5 runtime team. We not only shipped a great product, but we had a great time doing it.

Dates: A Coincidence … or not

It is quite interesting, then, to have yesterday, 10/19/2005, become my new proudest day at Microsoft. Exactly 5 years later …

You see … Yesterday we had a divisional meeting with Bill Gates to discuss the strategy of our division (that includes Speech, Natural Language, and the Assistance Platform). As part of that meeting, I had the pleasure of showing Bill what we’ve been working on for the last couple years for Windows Vista.

Yesterday was the first time Bill has seen Speech Recognition in Windows Vista running for real since we started the project. Bill was pleased with our progress. Quite pleased indeed. He didn’t think we could pull it off in the time we actually had to do it. But we did!

I’ll never forget yesterday —  10/19/2005.

It’s strange, then, that 10/19/2000 was so important to me as well. Exactly 5 years earlier… Hmmm… I wonder what that means is in store for me (and for speech recognition on 10/19/2010) 5 years from now.  That’s awfully close to 2011.

Is this all just a coincidence, or not?

We’ll have to wait and see…