MBIC First Day [General Session]

(Português) - (English)

This conference is similar to TechEd, where we have a list with breakout sessions, chalk-talks and labs. Then you choose what you want to see and shows up in the room to hear the speaker talking. During this day I'll participating in many sessions and will put here some of the interesting stuff. So let's begin at the beginning.

 

General Session

At the event opening some "mighty" Microsoft BI employees presented part of the Microsoft vision for the BI segment. I enjoyed seeing both presentations, coming from people recently hired by Microsoft (judging by what I saw, carefully selected). One thing that drove my attention was: both commented that during the interview they got impressed by the Microsoft's BI vision, so If two veterans in the area (competition at that time) get amazed, we can expect a lot. Now some stuff that I found interesting:

* BI Democratization.
If we ask in a company who benefits from current BI systems, we'll have an answer around 10% and 20% of the employees. If we ask how many could be beneficed from the BI, the answer will vary between 90% and 95%. Why this difference is so big? We need to go thru a process of democratization of the business intelligence, then it will be able to help all, not just managers and directors.

Projects Kilimanjaro e Madison (codename): the first is devoted to manipulation of high volumes of data (millions of giga/terabytes) and integration with DatAllegro. In a demonstration was possible to see many machines working independently over a big partitioned amount of data, answering a gigantic query in seconds (note: there was a instruction to clear the cache before execution). Scale-out, share nothing architecture, beautiful.
The second is focused in analysis and reporting self-service (going beyond what we have today with report models, for instance), that will empower the user (what we call people ready BI). Another thing that drove my attention was the Gemini (part of Kilimanjaro), where more than 23 million records where manipulated in memory (quick, very quick) using a "conventional" machine (with 8GB of RAM), but yet possible in a few years.

What I think is most important when we talk about BI at Microsoft is the value of the whole platform. We see Excel, SQL Server, Sharepoint and Performance point, working together in a integrated way and generating an huge value for those who explore all their potential.

More information? There is more people blogging... https://sqlblog.com/blogs/marco_russo/archive/2008/10/06/microsoft-bi-conference-2008-announcements-on-stage.aspx

A conferência é semelhante ao nosso TechEd, onde temos uma grade de palestras, mesas redondas e laboratórios. Aí você escolhe o que quer assistir e aparece na sala para ouvir o palestrante. Durante o dia de hoje eu vou participar de várias sessões e colocar aqui algumas coisa que eu acho interessante, então vamos começar do início.

General session

Na abertura do evento o pessoal do alto-escalão da área apresentou parte da visão da Microsoft para o segmento de BI. Gostei das duas apresentações, vindo de pessoas que foram contratadas pela Microsoft (a julgar pelo que vi, foram escolhidos a dedo) e se juntaram ao time não muito tempo atrás. Uma coisa que me chamou a atenção foi: ambas comentaram que durante as entrevistas ficaram impressionados com o que a Microsoft têm como visão para BI, então acredito que se veteranos da área (concorrentes na época) ficaram impressionados, podemos esperar muita coisa boa. Algumas coisas que achei interessante compartilhar...

* Democratização do BI.
Se perguntarmos a uma empresa quem se beneficia dos sistemas de BI atuais, vamos ter como resposta entre 10% e 20% dos funcionários. Se perguntarmos quantos poderiam se beneficiar de BI, a resposta será entre 90% e 95%. Porque existe essa diferença tão grande? Precisamos passar por um processo de democratização da inteligência do negócio, para que possa ajudar todos e não somente os diretores e gerentes.

Projetos Kilimanjaro e Madison (codinomes): o primeiro voltado para a manipulação de volumes gigantescos de dados (milhares de giga/terabytes) e integração com a recém aquisição da Microsoft (DatAllegro). Em uma demonstração foi possível ver diversas máquinas trabalhando independentemente sobre uma massa de dados particionada, respondendo uma consulta gigantesca em segundos (nota: a instrução mandava limpar todo o cache antes de ser executada). Scale-out e arquitetura share-nothing funcionando, lindo!
O segundo focado na análise e relatórios self-service (indo além do que já temos hoje com os report models, por exemplo), que irá dar mais poder para o usuário (o que chamamos people ready BI - pessoas prontas para BI). Outra coisa que deixou todo mundo de queixo caído foi a apresentação do Gemini (parte do Kilimanjaro), onde mais de 23 milhões de registros eram manipulados em memória (rápido, muito rápido) usando uma máquina "convencional" (com 8GB de RAM), mas ainda assim factível dentro de alguns anos.

O que eu acho mais importante quando falamos de BI na Microsoft, é o valor da Plataforma inteira. Vemos o Excel, SQL Server, Sharepoint e o Performance Point, trabalhando de forma integrada e gerando um imenso valor para quem utiliza todo seu potencial.

Mais informações? Já tem gente blogando... https://sqlblog.com/blogs/marco_russo/archive/2008/10/06/microsoft-bi-conference-2008-announcements-on-stage.aspx

[]s
Luciano Caixeta Moreira
luciano.moreira@microsoft.com
===============================================
This post is provided "AS IS" and confers no right
===============================================