Pros

In the technical/scientific side, the intelligence and friendliness of most of the co-workers. Hiring for good personality (lack of aggressive arrogance) and high ability. Willingness to help and share.

Few egotistical mainpulators I've come in contact with on my side.

Science + tech types who get their job done well.

Cons

Internal computer support for scientific needs is woefully underdeveloped (hardware and software) far worse than even basic academic labs, much less a Yahoo or Google. Essentially they have no clue and no distinction between some random app-server and hard-core scientific cluster simulation. Continuous downscaling about what IT will support or offer or install, enforced centralization, and of course totally forbidding science or devel to administer their own systems & software installation (i.e. be root).

See "advice to senior management".

Advice to Senior Management

CEO talks about how analytics are 'core' of company (indeed without modeling nobody would buy the mediocre software, except maybe Blaze---and that was acquired externally), but it still seems to be viewed --- or at least acted upon --- from the very top as just another random software company, which it isn't.

Major new product upgrades/releases get designed and decisions made with hardly any of the scientific staff knowing what is going in or having any input, and as a result, most decision end up being made by software engineering, with all change driven around their needs. Modeling is often just a

Where is the Chief Analytics Officer? Somebody with fundamental machine learning/modeling knowledge and the clout to drive long-term fundamental progress both from top-down ideas, and sponsoring individual scientist ideas (the "entry level" science is pretty good) to go from idea to product.

Fundamentally, scientists seem to be ignored by the rest of FIC management / business except when there's a bug (either FIC's fault or client's fault) and clients are whining. They just plug models into the small, obsolete, box set up for them by the software development design.

There is no communication from the FIC business side---most of the client partners seem to be inept in interaction and absent otherwise. Scientists don't even know whom to talk to about any specific client, and don't even know whom to ask. Many data problems come from client misunderstanding, but there is no push from FIC business to convince them to fix problems (as it would help the models they get).

There is no communication whatsover about what features competitors have and that we might think about working/improving on---indeed no communication about what the names of the the competitors even are, except through informal grapevine and employee departures.