[python] výjimečná práce na výjimečném security produktu - ThreatMark AFS

[Milan Čapoun]

Dobrý den,

dovoluji si vás oslovit - do naší úspěšné startupové společnosti ThreatMark
( <http://threatmark.com/> threatmark.com) hledám šikovné vývojáře Python s
tahem na bránu. Vyvinuli jsme produkt v oblasti Financial Fraud detection,
primární zákazníci jsou banky a finanční společnosti. Nyní se rozšiřujeme, a
hledáme posily pro náš team.

Pokud by vás níže uvedené informace zaujaly, neváhejte mne kontaktovat na
kontaktech uvedených na konci emailu.

 

Krátká popisná videa můžou přiblížit produkt rychleji, než dlouhé povídání:

 <https://youtu.be/2ZvlfpfsGlA> https://youtu.be/2ZvlfpfsGlA

 <https://www.youtube.com/watch?v=k-5xjoDXll8>
https://www.youtube.com/watch?v=k-5xjoDXll8

 

Threatmark je unikátní startup, kde je možnost pracovat na strategickém a do
značné míry revolučním produktu. Na rozdíl od velkých korporátních
společností jsme menší semknutý team s velkou mírou svobody v práci a hlavně
s bezprostředním výsledkem své práce, který je okamžitě nasazován do
produktu k zákazníkům. Hledáte-li větší naplnění než "chození do práce,"
případně vám nevyhovuje velké korporátní prostředí, může být Threatmark
místo pro vás.

 

V případě zájmu vám rád napíšu více a nebo nejlépe zavolám. V tuto chvíli
uvádím aspoň základní popis naší společnosti/produktu od našeho
technologického tahouna Kryštofa:

 

We are IT security startup. We are group of about 15 people based in Brno
(with second office in Prague).

Our goal is to secure web & mobile channels for internet banking
applications (and cards in near future).

Problems we are fighting:

- user installs financial malware (willingly as bundled into another app, by
exploit, by social engineering)

- users credentials get stolen (phishing, social engineering)

- user losts his credit card

- user's phone used for second factor of authentication was hacked/stolen.

 

We solve this by monitoring everything which happens on users device and
banking backend eg.:

- Transactions (by HTTP parser)(unsupervised anomaly detection / single
class classification)

- Navigation between pages in IB (Javascript / HTTP) (Markov chain stuff)

- Mouse movements and keystroke dynamics (Javascript) (features like fly and
dwell time, neural nets, random forests)

- Device unique fingerprint (Javascript) (entropy of features, computation
of distance of two devices)

- Accelerometer, gyroscope, GPS, touchscreen (hardware deviations caused by
process of "chip creation", user biometry)

 

There are many competitors on market, but they are focusing on single area -
e.g: how to technically detect malware, how to detect phishing.

Our approach is different, because we collect lot of data and from that we
are answering questions:

 

- Is it the same device as this user used before (or attacker) ?

- Is the same guy behind the device (behaviour biometry)

- Does the actions performed in session make sense (like sending 100% of
account balance to russian account)

 

We achieve this by combination of tech/security/hacker knowhow and machine
learning.

By this I mean that we employ both rule based approach, supervised and
unsupervised learning.

Some algorithms/techniques we use: neural nets, random forests, entropy
enomaly detection, bayesian networks, LSAD

 

Technology: python, sklearn, scipy, pandas, numpy, javascript (real ECMA
stuff, not jQuery animations :)), C++ (hunting milliseconds here),
Apache/Nginx/Haproxy, Unix, Oracle DB, MariaDB (hundreds of millions rows in
each table).

 

Project status:

We are there for about 2 years, received seed investment (about $200k) and
we are in the process of gaining another 1.5mil $. We have few PoCs and
first two paying customers. Our next step is to raise capital to accelerate
both development and sales effort. So far we are existing only in Czech and
Slovak republic, but we want to go "outside" soon.

 

We are not looking for regular employees/workers but for smart, excited,
enthusiastic winners with inner motivation to change the world.

If you are interested, let me know and let's discuss it further.

 

-- 

Milan Čapoun

Head of development

 

ThreatMark s.r.o. | Digital Identity Sensing Company

Pražákova 10| 619 00 Brno | Czech Republic

Mobile: +420 601 350 975 |  <http://www.threatmark.com/> www.threatmark.com

Email:  <mailto:milan.capoun na threatmark.com> milan.capoun na threatmark.com

 

Please consider the environment before printing this email

 

 

 

 

------------- další část ---------------
HTML příloha byla odstraněna...
URL: <http://www.py.cz/pipermail/python/attachments/20170221/57bc8a4c/attachment.html>