Roman's miles

IT Regional Architect Conference. Feedback from Atlanta.

As I announced in the past I've attended the conference in Atlanta. It was the first event of a series in US and the organizers plan to have it annually in the same format. Below is my impression.

First off, it was a great success! To my knowledge it's the only conference which is vendor or technology agnostic. Not only it is cross-platform and cross-technology it's focused on the architecture issues and built for architects. Not developers or sysadmins but enterprise, infrastructure, and software architects.

Secondly, for a local conference (by the way, participation cost is just $500 which is absolutely exceptional for such events) there were about 250 participants that again demonstrates success of the first event. It was quite an efficient networking event to rub shoulders with local folks.

Thirdly, the keynotes speakers were real stars and they rocked! Angela, Mike, Scott, and Rick all presented a very valuable topics and were great speakers. A very sound composition of the speaker - another complement to the organizers.

Lastly, I advise you to go to the next event in Atlanta or other cities and wish IASA to rise the bar next time bringing new speakers and more participants.

Utilizing your day. Rising the bar.

Do you constantly complain about the lack of time? Does your schedule slipper during a day as time goes and you miss some important stuff planned for the PM? It happens often to me and the missing part in my life is often running and workouts.
The always working recipe for me was getting to bed later, trying to V more complete tasks on the TODO list. Often, though, you can't make much at late hours and mind- or muscle-intensive exercise become unachievable at late hours even if you have a few before the day guillotine cuts your productive "today" from "tomorrow".

Ilia, my friend, taught me to swap the day edges for shoving planning and to make the most important (time-sensitive, risky-to-slippery, easy-to-defer, forgettable, etc.) very early. He wakes up at 5am, has first his swimming and workouts, and then starts the day. Very hard for me to do but I'm all envy.

Today my classmate, Kevin, told me he wakes up daily at 4.15am(!), has his run and workout, and starts (continues?) his working day at about 6am. Not that this fact would help me to wake up tomorrow but I'm sure such stories are exciting, motivating, and aspirational. Kevin answered my question "why do you do it?" with a very simple statement - "because I'm an person of achievements and love to get things done". Surely a motivating statement. And he's been doing it for already 6 years...

PMBA. Reader's questions

A reader sent to me a few questions about PMBA. I'll answer here to (hopefully) benefit future readers. An overall disclaimer is the answers reflect my own conclusions and opinion as of today. Of course it's questionable and may change in the future.

1. Do you know of anyone who have graduated from this program getting significant pay raise at their current company or a new company with a new job? I think that the only way to get a significant rise is to change the job one's doing. Apparently if it's just the next step on the same career ladder in the same profession the first raise won't be significant. If, on the other hand, one's changing the profession then even losing the formal grade may bring the raise in a long-term perspective. I personally ain't changing the industry I'm in and don't expect immediate raise (hope my boss doesn't read it:-) I believe in a long-run the MBA will make me more attractive candidate for executive roles than I am today and will help me perform better with my current job.
2. Will this enable you to rise through the corporate ranks quickly? Here I believe it all depends on the industry and the employer. In IT an MBA doesn't matter much. It may be a NO-filter for exec roles but if one adds MBA to the VC it barely affects her position. For sure not in SAP, for sure not for the current position. I'm clueless about other companies or industries although from conversations in the class I feel my perseption is quite common.
3. Will this PMBA help if you want to start your own business? That's an interesting question. I'm sure it will but at the same time it's not a must - many great entrepreneurs have reached great success with no formal degree. I'm saying an MBA exposes you to the areas you didn't know and makes you consider options you weren't aware of. Lots of my classmates are planning to take the entrepreneurship concentration.
4. Are any good companies contacting the career center to interview graduates of this program? I think all of the contacting companies are great. Whether you want to work there or not is another question:-) There is the career service in the school and we can contact personal career advisers and coaches. I haven't done it yet (just finished the second semester) but once I do I'll write about it.

To add I'd advise to read my thoughts before I joined the MBA and definitely attend open days in GSU and other schools you consider. I'm very positive also that before you go to school you must know why you do it, what your target is, and what you will do the first day after you graduate. Having answers will help you to find right school and justify your decision to pursue it or reject.

Travel plans

My (confirmed) travel plans for my friends and readers trying to catch me:

8/8 - Palo Alto
8/9 - Phoenix
8/14-15 - New York
8/22-24 Walldorf
8/25-29 Tel Aviv
September - vacation from travels due to expected Go-Live of our baby
10/1-4 - Las Vegas
10/8-9 Washington DC
10/17-20 - Munich
And probably some other unexpected trips so catch me!

Come to a "IT Architect Regional Conference 2007"

International Association of Software Architects hosts a series of IT Architect Regional Conferences at Atlanta, GA, San Diego, CA, and Washington, D.C on Sept 13-14, Oct 15-16, and Oct 8-9 correspondingly. The organizers' idea is instead of having one expensive national event leading to substantial travel expenses to host the conference three times in the country attracting much more visitors. Scott Amber gives the key notes on the conferences which makes it automatically as of a serious caliber.

You can click on the charter of your geographical preferences above and check the agenda. Many great speakers from the industry (software houses, consultants, and customers) present at the event and you'll find it difficult to pick up a topic of your interest from four concurrent tracks.

I'm presenting MDM at Atlanta and the capital - MDM "MDM - where business meets IT". I'm planning to talk about the problems MDM comes to resolve, why it's vital for today's business to address them, and what make an enterprise want to buy a ready solution instead of developing an in-house one. I'm planning to split marchitecture and tarchitecture as 1:5 (at least) so the geeks in the audience should not get too bored with ROI, TCO, and other exciting stuff.

Come to listen lead industry architects, learn something new, and rub shoulders with you colleagues. It's gonna be fun.

What am I doing at SAP?

I found myself answering this question many times introducing my group and myself in real life; recently I've been getting many emails asking the same question - so I thought why not to blog about my team and my role at SAP?

I work in Solution Office of the New Product Introduction group. It's a global group presented in North America, Europe, Israel, and Asia Pacific. There are just a couple of dozens folks in the team but all came with a sound background in ABAP, Java, or .NET and brought many years of experience in a variety of technologies.

The mission of the group is told  in its name. When SAP introduces new products (not just new releases of the existing ones) it may take some time until all the major divisions inside the company get up to speed with the products: we need to iterate the products through a series of adjustments accommodating cusotmers' needs and consummating solutions, educate project and sales forces, enable the ecosystem, and gain mind and market share. I don't want to call this phase "crossing the chasm" but it's a very critical period for a new product and while it does its first steps outside of the labs the product needs a team fostering and supporting it. This is a team behaving like a task-force unit. Again bringing a military analogy such a mobile group reacting in an agile and sensitive mode isn't burdened by revenue, utilization, or any other rigid KPI and can do whatever is needed for a product's success (comparing to SAP's armies of consulting, education services, sales forces, etc. using an approach of scalability and targeting revenue generation).

I'm leading a team of architects and experts in the Solution Office driving MDM - Master Data Management. Our mission is to provide customer success on early project implementations, enable internal project forces and the ecosystem of partners with knowledge and experience gained in the projects, roll-in  the feedback to the R&D, assure solution awareness among our customers and partners, develop solutions on top of the product, and do everything is needed to warrant product success. We closely work with other groups of the NPI virtual team (more than 15 different divisions), engage in pre-sales activities, bootstrap projects, assist in evaluating solutions, present the product at different conferences, evangelize the solutions on customer and partner events - so the team knows the product inside-out which helps us to identify the gaps, propose, and carry out solutions.

With only a handful of people in the team we've reached this year lots of things. We've worn a hat of consultant (proposing solutions for customer ), a solution architect (evaluating architectures with project teams), a trainer (preparing and teaching classes), an analyst (brainstorming with customers ways to resolve problems), a bizdev guy (creating  solutions on top of the product and educating partners), a sales man (pitching the solution in front of customers), and many others internally. The only our criterion for a task to be taken by the team is a company-wise impact. Our group doesn't exist to replace any other team although often a task can be dispatched to another team. At the same time usually there is no time, or enough expertise acquired by the the other team, or a shortage of resources, or they have other priorities. So we take the task on our shoulders, build a solution, and later - to achieve scalability and a massive impact - roll it out to an appropriate owner in the company.

We cherish (or boost?) a new product for about 12-18 months and when it reaches its maturity and simultaneously other teams acquire necessary experience the product departures from our radar giving in its place to the next one.

It's a very interesting group to work in with tons of dynamics and wide exposure to other teams in SAP. It's also a very unique position for touching non technical areas of our industry (such as marketing, business development, sales, consulting, training, etc.) without loosing contact with  our technical background. The only disadvantage one (but not we) may find is intensive travels. Some of us have spent about 100 nights outside of the homes this year and flown more than 100K miles but with the majority of international destinations it's hardly perceived as punishment:-)

SOA is around us

I've been running recently across a couple of discussions on what SOA is and thought it would make sense to re-post my old corporate blog on the topic here. This time I just change SOA's old name - ESA - to the new one.

A few days ago I went to my DMVS (department of motor vehicle) center to get my driver’s license. I went to the closest exam center, which is in Norcross, to take a written and hopefully a road test. For those who’ve never been there, the visitor’s area is a big hall with 20 counters, 3 big screens with numbers running on them, many seats and hundreds of visitors. I was told it serves between 400 and 700 people a day. Staying in the line hearing almost incessant announcments and looking at visitors and inspectors dashing about I had a firm sensation that all this very much reminded me something. Guess what?! – right, SOA! I was looking at every element of the interaction and more and more was becoming sure that this is quite a strong implementation of SOA, not in software but our real life. Let me explain.

As you get to the center a big guide sign meets you: “ALL visitors to the info desk!”. This is a Dispatcher-Controller pattern. The Dispatcher lady accepts all visitors and after giving basic instructions (“we accept only cash!”) and checking your docs (“where is your address proving bill?”) redirects you to a queue and assigns you a unique (ordinal) number. By the way, there are two-three ladies at the desk so there is scalability and availability in place! After you get your number you’re put in one of about a dozen virtual queues (or workflows) for your needs: passing tests, changing address, getting license back after revoking and so on. Every such purpose has its own workflow with special rules residing to it.

So each visitor becomes an “object” routing between different services according to a specific workflow and notified by proper messages. Such coordination is hidden from visitors but happens "behind the scenes" in a workflow managing system. This is the Dynamic Router pattern – the heart of the system. This Dynamic Router knows where to send you on the next step of your workflow depending on your results of the previous one. You’re advancing from service to service (from counter to counter) when your number is announced and it appears on those big screens I mentioned before. What about services? They’re supplied by brave DMVS employees standing at the 20 counters. They’re grouped by the functionality they supply and serve the next “object” in a service queue. For instance some fill out your form and verify the docs you brought. Others take your signature, fingerprints, and your photo. A third performs vision checks and so forth. Every service knows how to carry out its own operation. It neither cares where you came from nor does it know where you will go after. All it knows is to double check your “input parameters” perform its operation then “return some result” and finally “log” it to the system.

What is cool is you don’t worry about where to go next, either! The Dynamic Router service takes care about you. In practice you wait for your number to appear on the screens showing you the next counter to attend and announced by an electronic voice (“number A485 at counter 17”). You’re sitting and waiting. Now what about collision resolving? What if you lost your number? What if you missed your queue? In such a case you come to the Dispatcher lady which enqueues you again to the workflow hopefully to the point right after the last stage you successfully have passed. Big hint: messages can expire. A friend of mine, after successful passing her driving test, couldn’t wait to get the license and took all the documents home. When she came back a day after, the Dispatcher lady didn’t allow her to enter the queue on the last stage. Sorry, Sarit:-( So she had to start her workflow from scratch and thankfully she passed it again (now to the end of the workflow:-)

What about scalability availability and performance, you’ll ask? As I said every service is scalable over several counters. And hence available. Performance at Norcross was quite impressive (I came there at 10.30am successfully passed my both tests and at 1.30pm went out with my license in hands). Everything was functioning fine and fast. I like our projects implemented in a similar way:-) Not that I wish you to attend a DMVS center. But if you have to, pay attention to this brilliant SOA implementation and good luck with your tests!

How to read fast

Let me explain the title. I didn't mean how to read more words per minute that a bunch of different techniques claim to educate. My needs were how to increase the speed of comprehension of the read material, especially of business or studying books. Let's start with an extreme but clear case. Assume that we research a topic or read a book just to widen our horizon. If it's not our first book  in the realm the odds are good that we already know something about the matter; now we can either read every single word or skim through. The former is not efficient - we may read about the concepts we already know and waste our time. The latter is risky - we can miss something important and lose the book's value. Balancing two options is the kind of the fast reading I'm referring to in this blog.   

I already brought Brad's explanation on his way to read 8(!) books a week: "(a) no TV, (b) no kids, (c) four to six hours a night of reading, and (d) the willingness / ability to skim when things are dull". So the question is how to find an optimal pace for a book to trade off speed and comprehension most efficiently. I just started practicing this method and here are few my suggestions.

• Build expectations for the book. Before starting the book make it clear for yourself why you decided to read it. What is the question you're seeking answers for and why is this book indeed. Is it a particular idea, a set of examples and case-studies, or you need deep diving to the realm. Depending on the answer you may predefine what chapters to skim through. A good example of what I mean here is a pile of the books of Geoffrey Moore. Each of them is great but you don't want to read each of them from cover to cover.
• Always read the book's introductions (all of them).   In a read-every-word mode that was the only victim of my through-skimming. When I was cover-to-covering  a book I thought I didn't need to know what the book was about - I'd know it upon completion. In optimizing the pace very valuable information can be found in introductions. We need to figure out what the authors will be telling us on the next pages (or, at least, what they promise to) and adjust our expectations (see the first item). Often authors explain the book's structure that may also help us to make up our reading plan.
• Understand the book pattern. Every author is different but even every book of an author may be different. Still each book has its structure and style which even if not explained in the introduction nevertheless can be figured out after a few chapters. We need to recognize a question the author is about to answer, possible options for the answers, analysis of the dilemma if it exists, winning answers, and examples proving them. Depending on your needs you may decide what is "dull" for your particular case and accelerate reading. To make it precis the book's structure must be known.
• Take a sample and check the pace. After you answer the above questions you can start skimming. Usually, if you're aware of the problems the authors are explaining, an introductory part of a chapter can be skipped. A few paragraphs detailing importance and complexity of the situation are good candidates to save time if you know the area. Figure out where they get to the point and propose solutions to the problems. You may skip quite a few paragraphs and find that you miss the point. Don't despair. Come back but not to the very beginning. Take a half of the way and try again. After a few iterations you'll find the pace.
• A book may vary - re-adjust the pace. Rarely I find a book that keeps the same pace, style, and tone on all the way. It may happen as you go from chapter to chapter, especially when there are many authors. The pace should be adjusted and re-tested again. 

I hope you'll find the tips useful and waiting for your feedback. The final confession is I still read fiction books in a cover-to-cover mode (or just throw them away) and I'm happy when I get a business book which reading without skimming pleases me.

Update: I'm sure you'll find uesful "How to read a business book" by Slacker Manager

An escape door

Some people call it procrastination, others just laziness, thirds claim the motivation wasn't strong enough. It does all truly exist there but naming the problem doesn't help much in resolving it. The problem is how to advance you self in overcoming procrastination, attacking laziness, and fueling motivation towards the targets you've set up.

Each of us knows it from own experience. You set up a goal, start it with enthusiasm and passion then reach first blocks on the road. If the goal is well attractive you overcome it and drive farther but soon reach another obstacle on the way. You go over and keep on driving to the goal and new blocks appear and you strike back. But sooner or later you meet a point where the energy to gear is not as strong as desire to miss or escape. Your creativity in finding a valid, legitimate, solid reason to not to meet this obstacle becomes almighty indeed. A couple of examples.

Decided to loose weight and chose a program demanding not to eat after 7PM. The first day it was a pleasure to put to the test your willpower, the second day you still enjoyed your strength of mind and stuck to the program, the third day it was already tough: you hadn’t lost much weight yet but you’re very hungry! What’s a mistake to open the fridge at this moment! The stomach and the eyes fed the brain at a speed of a thousands thoughts per second by the reasons why you COULD catch a piece from the fridge.

Giving up smoking? Hadn’t smoked for already a week and got to a pub with your friends? All the receptors immediately started locating well known attributes preceding a good cigarette. Mix of smells, pub music, taste of alcohol, friendly environment – and you again found a million well-grounded reasons why one cigarette wouldn’t make you a smoker again.

My favorite example is early awakening. What’s unnecessarily creativity our drowsy brain displays in finding a thousand of meaningful reasons why it makes sense to put the alarm clock forward on an hour! All pragmatic plans set up previous evening are forgotten at that miserable instant and the big exciting target to become an early riser is literally in coma requesting an emergency treatment.

Not pretending to reveal a silver-bullet method I claim the problem in all the cases is discipline and finding a strong basis in reasoning not to give up. If you’re disciplined you don’t eat after the you diet doesn’t allow whatever tempting dish you see/smell. If you’re disciplined you don’t smoke however pleasant and comfortable it may be. If you’re disciplined you wake up a minute before the alarm clock is set. There are many ways to become disciplined and maintain the discipline on a high level. Here I want to present a good analogy helping me fighting giving-up a lot.

Imagine you’re taking a stair race. There is a hundred of floors in the building and of course participation is more important than victory. You start at the very bottom level at the parking with other runners in presents of reporters, friends, and supporters. A starting gun shoots and the crowd runs forward and up. You’re excited and vividly depicture your finish: there are your friends, colleagues, maybe your kids and your wife in the crowd of supporters; you’re exhausted bug happy, you’ve done what you wanted. But here you come back to reality and see it’s only 4^th floor – other 96 yet to go! You fight a pity feeling to your self and run farther. But passing 28^th floor the feeling supported by muscle pains and broken breath calls again and a tiny voice starts asking: why do you want to go so up, maybe if you could just skip off the 30^th stair-well it would be good enough? What a crazy idea was to go so high (and you’ve jut passed the 43th) - you clench the teeth and keep on. The voice becomes stronger and already is heard loudly – why? (floor #58). Your anti-reasons generator starts working in a full-power mode: maybe I’m not ready yet, maybe I need more daily trainings (that I’ll of course start tomorrow), maybe next time I’ll not start so fast, maybe I had to sleep more yesterday and had special diet for previous month. You’re seeking for an idea to justify giving-up (80^th floor) and here you see this – the escape door! What’s the luck! You’re rescued! You’re saved. You skip off this race and you’ve done it. But wait a second, there are no colleagues and friends on floor 80, it’s a silent corridor and nobody’s waiting for you there. You broke your plan, you worked hard but haven’t achieved the target and what’s most painful here is you and only you found this escape door. You didn’t believe you could go to the roof but were looking for a reason to escape. Once such a “reason” appeared [in your imagination or in reality] you immediately took occasion…

This analogy helps me to go to the end, not to give up, not to get out off the way. When I set up a goal I imagine a hundred floor stair race and visualize stair-wells. When I hear that tiny voice I don’t let it seize me and drive to the escape door. No escape door but the roof is the only meaningful target. Whatever creative thought comes to me when I’m in the middle of the race I ask my self: “isn’t it an escape door?” Every time I change my original plan to adjust it to new things I found on the way I check it very carefully not to get led astray to an escape door. I need the roof. I hope this analogy will help you to push back the destroying voice of doubt each of us hears when we’re on a mission.

How to make your startup a "cool to be working in" company?

I thought what could make people wanting to join and stay in a company even if a next door neighbor offers 20% more. I don't have a magic answer rather few candidate-like suggestions but before I go there let me tell you why I think it's important to retain your colleagues with you. The key factor is consistency and trust. People you can trust (not only in professional but more important in personal or even private  perspective) is the real edge in team building and the more such people work with you the more stronger this relationships and the team became. Your resource needs tomorrow may change but trusted people tha want to work with you will remain. Of course it means you have to "retain" only good people and ones you wouldn't take in your next start-up you should fire already today.

I think for a start-up it's better to have only real hackers in the team. I call a hacker one who would still continue playing with computers even if he wins a million dollars tomorrow. People who program not for living (or not only for living), people who don't leave their notebooks on evening and weekends in the office:-) A potential advantage of start-ups is due to a small number of needed people they can (theoretically) build the whole team from hackers. Joel promotes a similar idea providing numbers and reasons from his vast experience. So how to attract hackers to work for you? Here are few ideas without any preference order.

1. Office layout. Avoid cubicles! Just don't do it. So many friends of mine resisted to apply for big well known companies with one phrase - "donn' wanna work in cubicles!" Why, for God's sake, would you put  your best brain power into a 2x2 cell making them feeling like a factory cog? Layout your office with separate rooms for 2-3 people that like to have more privacy. Make spacious rooms for group work for people preferring collective environment - I mean seriously spacious rooms not meeting rooms remaining smoking lockups  in airports. Buy there a divan or comfort chairs, make people willing to stay there:-)
2. Office equipment. Buy comfortable working chairs. A chair is a working tool of a programmer, show them your respect their work (and over-work). Invest in cool boards (like this) or install cameras in meeting rooms to grasp and send board's drawing, buy phones with high quality speaker phones (for group discussions) and personal ear-phones. Buy a cool coffee machine!
3. Tools. Give only notebooks to the developers. Not a 17' screen 'game station but a good notebook for development. Undoubtedly it costs more than a desktop and hardly upgradable but for sure it will attract hackers (and retain too). Most important thing is if you hope they will think about your product evenings, nights, and weekends let them really do it. Give them home Internet for free.
4. Job duties. Let people deal with exciting things. I know it sounds strange but here is what I mean. People hate doing the same things (and those who like probably shouldn't be in your dream-team). Even if it was interesting in the beginning.  Give boring jobs to ones for whom it's still new. Write an installation? How exciting is it to create a new version for such a 'project'? Hire students for it and give somebody to lead them to try him in a position of a team or a project lead. To be a technical lead of the installation project is much more interesting/challenging/exciting than to customize it 15th time.
5. Horizontal promotions. People expect promotions and often leave the company seeking promotions. You can't promote everybody vertically - it's a bad idea to make all the programmers team leaders without a team. But you can give distinguished people new opportunities. An installation project lead is just one of the options. For people having strong technical experience offer to try the customer site - get involved into business aspects, interact with consumers, try to be on the front end. Give people opportunities to try themselves in the areas they probably never would apply for next job. An excellent QA engineer may try to take a small operation project and challenge a task the operations team lead has been dreaming of taking but didn't have resources. Such 'castles' may happen on a  temporary basis with part-time involvement. Isn't it a good bonus to let people a privilege to try something they may want for free?
6. Professional bonuses. Instead of paying money bonuses (or if you financial situation allows - besides that) let people think out a two week long project not related to your business directly and undertake it. If the team writes in Java let them write something in Ruby. Let them play with Linux if all running workstations are Windows. Let a HTML team create a prototype in Flesh. Let them study and then teach new technology. Let them be creative and offer projects that a) - interesting for them, b) - have more than zero probability be used in the future. Such professional bonuses may vary in conditions. For example prize #1 may be a project lead for full time, prize #2 - a team member with 50% involvement, #3 - 25%. At the end the project has to be presented to the entire team and the experience shared.

There are many ways to attract people and to make them willing working in your cop many even if the salary is not the highest. I put here my preliminary ideas came to me when I tried to answer to myself what would catch me in a company for a long period. What do you think should such a company has in its philosophy?