June 24th 2006 04:23 am

About Matt

I have been an independent BI consultant for many years and implemented numerous data warehouses and BI solutions for large companies. For the last 6 years I have been very busy writing an ETL tool called Kettle. This tool was open sourced in December 2005 and acquired by Pentaho Open Source BI early in 2006. As such I’m now Chief Data Integration for Pentaho mainly doing lead development for Kettle a.k.a. Pentaho Data Integration.

In our garden

42 Comments »

42 Responses to “About Matt”

  1. Alain.Debecker on 04 Jun 2007 at 21:50 #

    Hello Matt,
    I just did want to say hello. And thank you for the good time we spend in Berlin.

    And then… I find this blog.
    Just tell me. What kind of soft do you use ? I want the same.

    AlainD

  2. Matt Casters on 06 Jun 2007 at 17:58 #

    It’s called Wordpress, there is a link directly to your right.

  3. Lewis Cunningham on 23 Jun 2007 at 0:27 #

    Hi Matt.

    Thanks for the info on my blog. I got the opportunity to see some of the Pentaho stack at ODTUG this week. Very impressive. From the Oracle BI stack, I am most familiar with OWB so I will be starting with Kettle. The interface actually looks a lot like OWB but I guess there are only so many ways to visualize a transformation or mapping. ;-)

    Thanks,

    LewisC

  4. Matt Casters on 23 Jun 2007 at 13:00 #

    Hi Lewis,
    To a certain extent, Kettle was written out of frustration with OWB. As such I tried to do exactly the opposite of what OWB was doing :-)

    Be careful, you might even like Kettle. You wouldn’t be the first ACE that defects…

    All the best,

    Matt

  5. DaveMc on 19 Jul 2007 at 12:55 #

    Hey Matt, you’re a legend!

    Just getting started on Kettle, and it’s really, really good.

    Well done, all your hard work is well and truly appreciated.

  6. Ali Akkas on 12 Nov 2007 at 11:58 #

    Hi Matt,
    Congratulations. Kettle is a great tool for ETL and it is very easy to use and getting things done without too much learning curve. Thank you. Keep up with the good work.
    Regards,
    Ali Akkas
    Oxford, UK.

  7. Gerson Reis on 03 Dec 2007 at 21:11 #

    Hi Matt,
    I am Brazilian and newer in kettle and i’m tryng to create a transformation in spoon, export to xml and run this transformation in a java class, but I’m not getting success. I don’t know if this can be done. and if can how to do.
    I have read about this in the internet and pentaho forun but it’s not clear for me I need something more explained, like step by step.

    I have much dificult to find help about the kettle java api on the internet. if you can incate something.

    Thanks a Lot
    Gerson Luiz dos Reis

  8. Matt Casters on 03 Dec 2007 at 21:26 #

    Hi Gerson,

    I understand the temptation to ask questions on this blog, but it would be much better if you could turn to our forum for this: http://forums.pentaho.org/forumdisplay.php?f=69

    Thank you for your understanding!

    Matt

  9. Gerson Reis on 03 Dec 2007 at 21:56 #

    I’m Sorry for this post it was not so happy for me hehehe

    I tried to post in the forum but i can’t in the that time.
    I will post ther this question Ok.

    Thans and Sorry Again.

    Gerson Luiz dos Reis

  10. Edward Gibbons on 16 Jan 2008 at 9:59 #

    Thanks Matt,

    Your Kettle Integration software is truly spectacular. I was attempting to learn OWB until I found Pentaho Data Integration.

    I have since created many successful transformations and uninstalled OWB from our machines.

    If only all enterprise software were this easy.

    Regards,

    Edward Gibbons
    Southern California

  11. scott china on 25 Jan 2008 at 12:52 #

    Hello Matt,
    I just did want to say hello. And thank you for the good time we spend in china.

    scott

  12. Van Zyl Kruger on 26 Feb 2008 at 19:14 #

    Hi Matt,

    I have been using Kettle since its first open source release. I have implemented at many customers. Thanks!!!

    I have had this problem in the latest release. I do daily updates from an ODBC source database. After the process has run for a month or so the reading from the database really slows down (from 2000rps to 150rps). The only thing that helps is to reboot the server OR delete the previous ODBC config and re-config. Any ideas?

    Thanks again. BTW I am presenting Pentaho to the largest BI conference in Africa tomorrow.

    Regards,

    Van Zyl Kruger
    South Africa

  13. Matt Casters on 26 Feb 2008 at 22:34 #

    There is probably a memory leak in the ODBC driver somewhere (some DLL). If you can, try to use a direct JDBC connection.
    In the future, please post problems like this on our forum or create a bug tracker.
    Good luck with the conference!

    Matt

  14. Salah on 27 Feb 2008 at 20:35 #

    I can guess that Kettle is one of greatest open source data integration tools.
    I searched on internet for a comparsion between Business Objects data integrator and Kettle but couldn’t find.
    I need such comparsion in study to my company that is planning to implement a data integration tool.

    Any Help?

    Thanks.

  15. Matt Casters on 27 Feb 2008 at 21:28 #

    Salah, why don’t you download both tools and see for yourself?
    What’s that you say? You can’t download BODI? I guess that’s a big difference already, isn’t it.

    Matt

  16. Fred on 29 Feb 2008 at 2:00 #

    Matt,
    My company is in the beginning stages of implementing OWB. Could you give us some reasons why Kettle is better? Any information you could give might save us a lot of time and effort with OWB.

    Thanks
    Fred

  17. Matt Casters on 29 Feb 2008 at 2:06 #

    Ouch Fred, do yourself a favor and read this blog entry: http://www.ibridge.be/?p=65
    If against my advice you would still go for OWB, make sure you have someone with very good in-depth Oracle knowledge on your team.

    Matt

  18. Lilia Muñoz on 13 Jun 2008 at 9:24 #

    Hello Matt, thank you for your blog is very helpful. I am looking for the metamodel of the tool Kettle, I need help

  19. yxskkk on 15 Jul 2008 at 2:54 #

    Hi Matt,I am chinese,I want to learn kettle and make friend with you,am I?
    My msn is : yxskkk@263.net~hehe

  20. Nicolas Nakasone on 15 Jul 2008 at 19:17 #

    Matt, fantastic work…!!! My name is Nicolas Nakasone Bi Consultant too, in this moment for here yet don’t ear about open source, but with your innovate and join effort, the bi open source will conquest all the world.

    Best Regards from Lima, Perú.

  21. Babs on 28 Aug 2008 at 5:27 #

    Hi Matt,

    The blog is a great read, thanks!

    We are currently evaluating buying an ETL tool. What would you say to those (vendors) who say,

    1. Open source tools such as Pentaho, Talend etc are for smaller ETLs managing small volumes and are for small ETL jobs not for entrprise class ETL?

    2. If you have other tools from them in your environment for example, BOBJ BI and BOBJ Data services OR IBM DataStage and Cognos, then the consolidation and integration of meta data from these tools allows for easier management resulting in reduced workload and better quality of data.

    3. How does a tool like Kettle address this issue when an organization has a different BI platform?

    Thanks,
    Babs

  22. Matt Casters on 28 Aug 2008 at 9:02 #

    Hi Babs,

    at Pentaho we have been selling professional support and services for more than 2 years. In that period of time we’ve gathered a nice collection of customers. Every now and then we announce this over at Pentaho.com so go there if you want to have a look at a few customer cases.

    Answer 1. Now that Pentaho Data Integration offers performance equal to or better than the commercial vendors (you should try yourself!!) the only defense left for the vendors is FUD at the moment (Fear Uncertainty and Doubt). We just had a customer of ours do their own benchmark against BODI and they couldn’t find a situation where BODI was faster than PDI. (PDI was at least 20% faster) In the customer references we have a testimony of a company that said they replaced OWB and saw performance go way up as well. Mind you, in a lot of these cases, these companies would still have selected Pentaho Data Integration if it would have been 20-50% slower!

    Answer 2. Ironically, it’s not BOBJ, COGN, IBM nor INFA that are open in their specifications and metadata. Ask yourself this question: how is the inclusion of more closed software in your stacks going to improve transparency? Obviously, if you have a lot of money and don’t mind the perpetual vendor lock-in, it doesn’t matter. For a lot of organizations, completely open systems are the way forward.

    Answer 3. In large corporations with large deployments of proprietary data integration tools, the purchase and maintenance cost of the software is substantial. However, the invested cost in terms of time (work) is usually a lot more. As such, it becomes a huge vendor lock-in. For example, I’ve heard of a company that had their people work in shifts because they couldn’t afford any more DataStage workstation licenses. In the end, what it comes down to is that companies usually budget costs per project and that Kettle is then being deployed for one small separate project (usually to see how good it works) and then another, and another. In these configurations, it works alongside the proprietary tools. The fact that Pentaho Data Integration is very easy to set up, configure and manage has something to do with it I guess. The overhead of maintaining Kettle as an extra tool is far far less than the purchase cost of additional licenses or paying more maintenance costs. Additionally, in the long run (4 years or more) it gives these organizations hope for better times.

    Take care,
    Matt

  23. taoufiq on 17 Dec 2008 at 22:19 #

    Bonjour Matt,

    je veux juste te dire bravo pour ce belle outil Kettle vraiment je l’adore
    je suis entrain de convaincre mes supérieurs d’opter pour la solution open source Ketlle au lieu d’acheter l’autre

    juste une question si tu peux m’aider quels sont les choses clés que je dois absolument leurs parler pour les convaincre

    merci beaucoup

    Taoufiq

  24. Matt Casters on 17 Dec 2008 at 22:37 #

    Bonjour Taoufiq,

    Tu peut toujours m’envoyer un e-mail ou tu peut utiliser notre forum :

    http://forums.pentaho.org/forumdisplay.php?f=135

    A+,

    Matt

  25. Jihong Liu on 26 Feb 2009 at 21:23 #

    Hi Matt,
    Does Pentaho Data Integration supports transformation level transaction now?
    I could not find out this feature in version 3.1.0

    Thanks
    Jihong

  26. Matt Casters on 27 Feb 2009 at 1:50 #

    Sure it does Jihong (hint: “Unique connections” option in the transformation settings). However, please post your questions to our forum.
    Thank you for your understanding.

    Matt

  27. Terry on 09 Mar 2009 at 4:20 #

    Hi Matt,

    I happened to know kettle recently and realized that kettle is great tool.
    Currently I am working in the field of distributed computing.
    My interest is making distributed computing easy for users who are unfamilliar with
    I think Kettle is good solution for that purpose.
    I’m trying to develop hadoop(http://hadoop.apache.org/) components for Kettle because Hadoop is one of the most famous distributed computing plartforms.
    I will contact you again with some sample plugins.
    Any advice and comment is welcomed

    Thanks
    Terry

  28. Matt Casters on 09 Mar 2009 at 9:54 #

    Hi Thierry, for samples of plug-ins, you can visit the PDI Plugins page:

    http://wiki.pentaho.com/display/EAI/List+of+Available+Pentaho+Data+Integration+Plug-Ins

    Feel free to post more questions on our forum.

    All the best,
    Matt

  29. taoufiq on 28 May 2009 at 12:39 #

    Bonjour Matt,

    je me demande si il existe des certifications pour les utilisateurs de Kettle.

    puisque comme je suis intégrateur de PDI il me demande souvent si je suis certifié Kettle, puisque le marché parle plutôt langage certification que compétence.

    A+

  30. Bernie on 19 Nov 2009 at 19:48 #

    Hi Matt,
    Did you have a blog post that said, basically “I’m a former OWB expert who was sick of OWB, so I created Kettle.”?

    I remember reading that, but I haven’t been able to find it anywhere.

    Thanks

  31. Matt Casters on 19 Nov 2009 at 21:47 #

    Bernie, check the “Making the case for Kettle” post.
    I didn’t really put it as colorful as that but it was indeed very much like that.
    During first 4 months into the last project I did with OWB (9i) we got 6 (six) serious bugs *accepted* by Oracle. If you know how hard it is to get bugs accepted by Oracle, you know what I’m talking about :-)

  32. ahuoo on 16 Jan 2010 at 14:03 #

    ????????????????Matt? ??????? ?? wonderful tools , Matt , you are sooo cool !

  33. ahuoo on 16 Jan 2010 at 14:06 #

    o ,so sorry ,your blog have some problem for chinese

  34. Shaheed Fazal on 14 May 2010 at 10:41 #

    Hi Matt,

    I was wondering whether Kettle can be used for the scenario below:

    - I have a master list of products
    - I want to match another list of products (format and codes are not the same)

    I was thinking of using some sort of fuzzy matching but I want humans to verify each match because I am dealing in drug names and if one character is out then it can cause huge problems. Is this possible?

    Also, I get these lists daily so will Pentaho store the accepted matches in some sort of index?

    Looking forward to a favorable response.

    Shaheed

  35. Matt Casters on 14 May 2010 at 11:22 #

    Shaheed, you could use the “fuzzy match” step of PDI 4 in combination with some web logic, perhaps using the Pentaho BI server.

    Good luck,
    Matt

  36. Razane on 16 Jun 2010 at 12:21 #

    Hi Matt,

    I need your help, I have to extract the content of a binary object ,and I don’t know how to do it , I tried with talend but I failed.

    Thank You,

    Razane.

  37. Razane on 16 Jun 2010 at 12:25 #

    My question is : How do I extract a Blob from my database Oracle ?

    Thanks

  38. Matt Casters on 16 Jun 2010 at 12:48 #

    Razane, you can ask your questions on the Kettle forum over here:

    http://forums.pentaho.org/forumdisplay.php?f=135

    Good luck,
    Matt

  39. Leonardo Müller on 07 Jul 2010 at 17:00 #

    Hello Matt!

    I work with the Kettle for a year and a half in a large project in Brazil. We’re developing to the highest courts of the country with vast amounts of data and architecture and business rules too complicated. Caio Moreno Junior from Sao Paulo, who knows you, told me that perhaps if interessase Pentaho to use our project as a case due to its complexity. If you want more details write to my e-mail and keep in touch!

    Leo

  40. Mihai Manea on 16 Aug 2010 at 23:11 #

    Hello Matt
    I did an internship in one of Sybase’s branches, where i developed a Pentaho ETL plugin for bulk loading data from Sybase IQ database.Now I am back in the university and because i liked Pentaho i would like to join the Pentaho’s community as Java developer.
    Could you please give me some info about how I can bring my contribution and with which current community developers I could work.

    Best regards
    Mihai Manea

  41. Matt Casters on 17 Aug 2010 at 10:03 #

    Hi Mihai,

    Anyone can be a contributor. Simply create a JIRA case with your source code attached.
    If you want to contribute regularly then send me an email and we’ll set you up with write access to our repository.

    Thanks in advance for your help!

    Regards,
    Matt

  42. Mihai Manea on 19 Aug 2010 at 19:36 #

    Hello Matt
    Regarding your proposal on contributing regular, that it would be great!
    Could you please give me a contact email where i could reach you.

    Best regards
    Mihai Manea

Trackback URI | Comments RSS

Leave a Reply

Pentaho world image