Monday, November 23, 2009

Awefully long time no post

Gosh, it has been so long for no posting. It seems I have really graduated from MIT;). Time for a new blog link. So, here's it is: http://benjamin-jiang.blogspot.com/

Wednesday, April 8, 2009

Sponsoring MIT Sloan Sales Conference

nexiwave is now a technology sponsor of MIT Sloan Sales Conference. The Sloan Sales Conference is really an awesome event. Besides all the great sessions, just for the fact of networking with industry peers is a great opportunity.

nexiwave is giving out a special promo code for anyone who goes to MIT Sloan Sales Conference. I strongly encourage everyone to sign up for this event: http://www.sloansalesconference.com/

Sunday, March 29, 2009

Support from SDM Community...

I am really grateful for all the support that was received from the SDM community, both from alum and current. Thanks to Samir, Carrie, Cyndi and Shashi, of course. Really appreciate!

Of course, not to mention the support from SDM program as well.

Thursday, March 5, 2009

Made it to MIT 100K Semifinal

Great night after all. Our venture, nexiwave.com, made it to MIT 100K Semifinal (5 out of 35 in Web/IT track). It has been really hard work. It's nice to get validated in some sort of way.

On a side note, I would like to congrat SDM program too. SDM has 5 teams made to MIT 100K Semifinal. That's 5 teams out of 30 teams. 17%, wow, that's great record for SDM. Congrats to Alex, Chad, Jeff, Arthur and my partner Cynthia.

Monday, March 2, 2009

Help from MIT High Performance Computing group

Kurt Keville from MIT High Performance Computing group, who also happens to be SDM09, has been helping with the SDM cluster. I really feel fortunate that Kurt said he has a few GPUs and a few small clusters that can be hooked into the cluster. Hopefully, we can attach them to the cluster soon. It should improve the performance greatly. Thanks, Kurt!

Monday, February 23, 2009

Blogs and SDM

Today, Nisheeth grabbed and chatted about using SDM blogger to promote sdm. I think it's a superb idea. I researched Yoav's blogs extensively before decided to go SDM. Actually, it provided at least as much as information as sdm.mit.edu.

To my fellow sdm bloggers, this is what I am thinking:
  • Discuss our interest areas and have our blogs a little more specialized in those area: SDM Courses, SDM student life, MIT life...
  • Have sdm.mit.edu link our blog entries;
  • This should involve both present and past SDMers. Or anyone who wants to contribute back to sdm.
Nisheeth also mentioned monthly news letter to SDMers (sorry stealing your thunder a little here). That's a great idea too. We should propsoe this to Pat.

Let me know what you think...

MS Speech SDK and start-ups

My teammate of strategy class, Alex Stevenson, reminded me that there is a speech sdk available from MS. I played with it, but haven't found a way for speaker adaptation though. Will keep pushing on this front (when got time;)). Alex worked at MS before came to sloan and was on the direct2d team...

On another note, apologize for not post for so long. Too many things going on and didn't know what to post;). nexiwave is doing very good at this moment, with the first contract in hand and turning in profit-making company;-o. Other than that, I have been really working really hard: nexiwave is the main focus of course. A lot of small features, bug fixing, etc.

I am also involved with 3 more start-ups, with Barr and Jason mainly. I am glad that Barr is very interested in my facebook idea. Speaking of that, I need to do the slides for the facebook idea for the 100K entry. I am on a side-team with MBT and helping out with SENLive.

Too much to dump for one entry. Nexi time;).

Saturday, February 7, 2009

Broke my promise yesterday...

Leigh had an impromptu gathering at Tavern in the Square last night and I promised I'd go. I didn't make it though, but my excuse was I was with another friend. I am sure my fellow SDMers had fun there.

btw: rushed to SDM to install the two new and final rack servers around midnight. I realized I used up all available ports in the switch. In the need of a gigabit switch again.

SENLive

This afternoon, I had a good chat with with CEO of SENLive, Jason. We discussed how I can help out with SENLive.com. Very interesting problem to solve and we made a few very interesting differentiation points. SENLive definitely has a huge market to tackle, but also not with an uphill battle though. Jason is also interested in enrolling into MIT 100K. A lot of work is ahead for this.

Thursday, February 5, 2009

Spring already

Looks like Spring term is here already. I shopped for a few classes with one of the other fellow SDMer for some strategy class. Still deciding which one to take. Am I taking way too many strategy classes? Maybe I should sign up for some leadership classes instead. I do need one more class though.

The reason of being this late for the classes was because of the craziness of doing many things at the same time: a start-up to build, a family to care and ... a MIT Master degree, probably one of the most demanding University in the world;), to finish. Life is certainly the way I wanted it to be: don't want to know what for sure will happen six month from now: It seems I am slated to graduate this July, but we still haven't decided where to live yet. Oh, the craziness.

Monday, January 26, 2009

What a difference...

I had problem to find a home for the two rack servers that I bought. They are extremely cheap, but also extremely noisy. Brian Bulmer, the IT support for SDM/LFM program who had been really kind in support my effort, could only take them 20 minutes. Even for that, I gave him credit. Those two rack server are real beast. I looked up the spec for the 10 fans in the two servers. All of them are max air flow. As you might have guessed, max air flow also translates into max noise. Anyway, it has been bothering me for not able to find a home for these two servers. The reason I wanted to use these two servers was the price. Wayne Liu from SDM07 was kindly enough to offer to build machines with the current market specs. However, that would cost a lot for a single machine. For the price of today's single machine, I could get 3 or 4 racks servers. Consider we probably need 5 machines for redundancy. The cost saving is just too big to resist.

Anyway, Dave Schultz from SDM program office was very kind to let me use SDM storage room for the rack servers. There I have it, my own server room. I probably can dump 42U in there without bother anyone;). Thanks, Dave.

There is no network drop in the storage room and I had to get some wire thru the door and used a Linksys switch to share the one network connection.

After the 4 beer Chinese New Year celebration at Jeff place this evening, I figured I might as well move all the machines to the room. I did and it turned out to be a mistake. The network speed is much much slower. For the same page on nexiwave.com, from the room, it might take 3-4 seconds. If I connect it to the wall directly, it only took <0.5 second. I am thinking the Linksys is the pulpit. Anyway, I had to un-do the move. Now, the network speed is perfect again. I am thinking buying a cisco switch to replace the linksys, or... jab;)...

Tuesday, January 20, 2009

Google's Crawler...

Looks like Google's crawler has finally decided to crawl nexiwave.com. Maybe I should sign up hubspot to get more traffic from google....

C:\>nslookup 66.249.73.129
Server: STRAWB.MIT.EDU
Address: 18.71.0.151

Name: crawl-66-249-73-129.googlebot.com
Address: 66.249.73.129

Wednesday, January 14, 2009

First batch of SDM invites is out

what a holiday I had. Worked like (replace with any of your favorite curse word), a lot. Anyway, just sent out the first batch of a small group of SDMers to register the at our website for conf calls. Finally I have some time to blog a bit.

Thanks, Sergey, for being the first one to sign up.

let's see how it goes.

Friday, December 19, 2008

Signed up Comcast Extreme/Ultra Cable Internet

I was upset because there is no FiOS in my apartment building. Today, I noticed Comcast offers an Extreme edition, which is quite comparable to FiOS: 50MB/s download + 20MB/s upload. Both with a whopping price of >$140/month. The good thing about Comcast is that you can scale up and down at any time. It's pro-rated. No need to wait for a month. Of course, FiOS claims that they are the true fiber optical all the way, it will be interesting to see how fast Comcast extreme edition is and if it's constant. Sorry, Sergey, will support verizon when they are here.

btw: Nickolay, who is in Moscow, told me that he only pays $50 for that speed in Moscow. It made me wondering a bit.

Shall receive the new modem soon.

Saturday, December 13, 2008

CrossBrowserTesting.com

I found a great site for testing cross browser testing. Naturally, it's called as the title of the post says. They give a 5 minutes free testing window and a super-duper sign up window. I tested the site we are working with various browsers. It seems IE is always the PITA.

Great site. Recommended.

Thursday, December 11, 2008

Congrats to Dan

Congrats to Dan. Linda and Dan got a consulting gig for a community college out in the west. Mostly amazingly, they mostly used what they learn in System Arch, SD stuff to win the contract. I guess this concludes a little to this Sergey's post. This stuff can be useful in some situations. I'd still have to say, in some remote situation, such as the west ;).

It seems they can get busy in the next couple of months and most likely to September.

ESD.57 and Disruptive Tech

(started writing this on Monday)
Even though there is still one more class next week, today's like the last class for me for ESD.57 since my final presentation was today. Understandably, since I am into this Voice thing, my topic is Voice 2.0. Well, this is kind of late topic for this class. My original topic was Mobile Broadband and whether it can be disruptive technology to AT&T. As we went on, I realized that Mobile Broadband is not going to be disruptive technology for AT&T. Well, it can be disruptive to AT&T if it wasn't implemented well. It is not the Clay Christensen's definition of Disruptive Technology.

This is my understanding of Disruptive Technology: a technology that
1. serves a new market. This new market must have no interests to existing players, either because their existing customers have no interests
2. Or, serves an existing, and possibley low-end market, that the existing players have no interests. The reason can be : low margin,

The true meaning of be disrupotive, I believe, is serving a market that the existing players does not want to serve (new market), and even eager to get away (low-end).

For a new market, this can because the existing customers of existing players have no interests to the tech at that moment, such as the famous disk drive case. However, this does not mean the customers will be interested in the future after new use was proven to them.

For low-end market, this can be because the very low margin of the low end, such as the famous steel mill case. The very nature of low end focus will keep drive up and away of existing players from the upper and upper market.

Or, I would say a truely disruptive tech is something that disrupts existing players in an unprepared way. In the case of AT&T and Mobile Broadband, everyone knows that's the future of telco and AT&T knows what that means to them. It is paying huge attention to it, Mostly importantly, this is also a natural extension to the existing AT&T network. It is something they are eager to get into. Mobile Broadband is an incremental innovation.

Ok, let me go back to voice2.0. For the low end, I presented JitterBug and MOSH Mobile. JitterBug's focus on senior market, a market that is not AT&T's focus. The very high operating cost (US based live call center) is also not attractive to AT&T. For MASH Mobile, its ads-sponsored free mobile service is for sure not AT&T's interest. So, how will they disrupt AT&T? It can happen in a few different ways. JitterBug could take away the whole senior market. It can then move up the value chain and focus on middle-age market. The Live Human base call center can be attractive to the middle age as well. No more fighting against the phones or all the so-called smart features (which is not truely AI).

For the new market, I promoted the future focus of the content that is actually flowing through the voice channel. There are a few reasons that why it can happen now, such as speech recognition technology, storage cost, etc. In terms of market to explore, I suggested to look into the lead users. What are their pains and how they are solving it right now.

Tuesday, December 2, 2008

Final Presentation for ESD.36

Hooray, ESD.36 is over for me. We got our presentation just in time (eg, this morning 10:30). The term project is a bit challenging for me for many reasons. First, I am not in particular interested in manufacturing, even true for Aircraft manufacturing. Second, both my team members are distance. Thuy got a new job yesterday and Tarek is on Beirut. Last, the schedule was super tight. We started really working on the ppt just a couple of days ago. I wasn't even sure if we can get all the slides in time.

In the end, we delivered. I am so glad that ESD.36 is done.

In retrospect, it wasn't a bad class. I think I picked up some useful stuff, such as CPM, DSM and many other tools and methods. I doubt all of them will be useful, but I think it's worth to know those tools and, more importantly, terms.

Sunday, November 23, 2008

met with Shashi

ok, this is just some doodling from my chat with Shashi last night, but some very, very important advices:
1. get the first check, don't be shy to ask for it. If you are doing work for them, you deserve to ask for it. Focus on the market.
2. Try to get a deal with voip providers and work on those leads.
3. offer something for free and charge for premium.
4. You will need someone on board full-time to get any creditability.

Man, friend is always important.

major milestone soon

We finally done with the speech training with very acceptable result. More training and improvements are obviously on the way. Training is really time consuming, every iteration takes 3-4 weeks to finish. This is important and encouraging. We will have a demo-able system in the matters of days.