Virtuous cycle

Bartlomiej Owczarek weblog

You are currently browsing the Virtuous cycle weblog archives for February 2008.

Return to 'Virtuous cycle' home page

Inteliwise, Polish AI company, will go to NewConnect

Internet Standard published interview with CEO of Inteliwise, a Polish company which deals with AI technologies in search. On market since 2005.

The company plans to debut on NewConnect, a Polish stock exchange for high-tech companies.

I don’t have time to investigate it’s products at the moment, but the Company seems to understand what is most important for the AI:

Inteliwise avatar

A decent front-end.

Other thoughts: looks like an interesting company to watch (or even invest in).




Poland reduced unemployment rate faster than anyone

This according to Gazeta Prawna, which quotes Jan Rutkowski, economist at World Bank.

Unemployment fell from 20.7% in 2004 to 8.5% lately. Spain, apparently, took 20 years to lower their high jobless rate.

I made a nice chart about labor market once, maybe I should update it.




Freakangels, online cartoon

Found freakangels today. It is published weekly. So far just two episodes.
20080223 freakangels.jpg

I think I like it.

P2P lending on bootstrap – notes

Between burger and an ice cream in Hard Rock Cafe I jotted down notes from today’s bootstrap. I needed them in electronic form anyway. I wrote down stuff which seemed useful for me, so it’s not guaranteed to be complete.

This time it was in an unusual for boostrap form of a panel. Meeting took 3 hours, but I found it very interesting. Technical people might beg to differ, though, as tax/legal discussion was really exhaustive (exhausting).

I made some comments below in the notes, but here I can also share my “first impression” of the P2P lenders present. It is really first impression because I didn’t bother to check them myself, so some doubts below might be result of this.

  • Finansowo – approach to start as simple as possible, no collections or support in executing transactions between peers. The biggest risk is that some things that they ignore seem critical (collections).

  • Monetto (blog)- most advanced approach, ensured partners from the beginning, including bank (mBank) and collectors. The biggest risk is that they start with a very complicated machine that no-one will use.

  • Kokos – midway between the two, with the advantage that it is already up and running.

On a first glance I like Finansowo and Monetto, because they follow clear-cut concepts of simplicity and exhaustiveness, while Kokos was less clear, at least as much can be judged from the presentation, and also takes cautious approach regarding some legal challenges (e.g. possibility to grant anonymity and stick to electronic contract form).

(read more…)

Uh oh

I lost the debit card in Kiev. We were in hurry for the airport. And unlike ATMs in Poland, in Ukraine they give money first, card second.

Not that Polish setup is bulletproof, either. Some 3 years ago I forgot the money.

Clustering

I have this rather large file with loan data and I’m playing some statistics on it. During the day it was standard pivoting. But now in the evening I decided to check whether I can get anything by a fancy black-box clustering.

The best thing to get, of course, would be a set of clusters with significantly different loan performance (i.e. share of bad loans) that the others.

I’m using Cluto.

It is not particularly user-friendly. It requires input files that I need to kind of manually generate from Access and then fine-tune. And it is not excel add-in, but a command line program. But thanks to this it can handle my 100k records (my Excel version has 64k rows limit).

So far no results. But wait… just finished computing using the graph method. It took 17 minutes.

Nope. At the moment most distinctive cluster is ca. 6.5% better that the average in case of defaults. And ca. 11% in case of defaults considered fraud. I’m not impressed.

Guess they need to give me more data from the application. Currently I test on 6 variables and some are loan and not customer related so there is a field for improvement.

Or maybe I should read the manual some more and figure what what are the different optimization methods.

Bromba on the road

There are couple of pending things to mention and one of them is that my noble colleague Wojtek, at the moment also known as bromba, is currently carelessly traveling around the world. Together with his beloved Kate, at the moment also known as a small mongol (??).

They have a blog (but in Polish) and post pictures from time to time and in general make me think that traveling around the world might not be a bad idea.

Best SF movie this year

I just saw Cloverfield. By Bad Robot Productions.

Bad Robot

I know that the year is still young.

Microsoft to take over Yahoo!, aim for more failure

Strange, but sometimes things that are expected to happen, actually do happen. Microsoft announced a hostile bid for Yahoo.

I so like the bid to succeed. That’s because I feel strongly about the result of this take-over and I would love to verify my hypothesis.

My hypothesis is that the market shares of the two do not sum up. MSFT will lose focus and instead get entangled in restructuring of Yahoo!. Users will get the combination of the two worst, in my personal view, experiences – Google will be able to differentiate even more strongly through simplicity and focus, playing contrast to the new behemoth with combined dna of committee-driven software factory and a portal.

Or maybe the above will be proven wrong and Google will at last get a worthy competitor.

(Even though falling price of MSFT stock indicates that I’m not alone in my opinion of the contrary.)

In any case, this merger going ahead will be decisive, this way or another, for both Google and Microsoft.