Business Intelligence/Big Data as workfield - No experience/qualifications

Yogo

You guys are a helpful bunch and I hope you can shed some light over my situation.

Here's the story:

I am educated as a video game researcher (M.Sc. Games) with a Bachelor background in English in Denmark. Because I am a researcher and not designer or developer, I never learned nor was taught how to code at my university. I've tried teaching myself, but it has proved to be harder than just so. It is doable, but it will take considerable time before I have the basics down.

Since graduating I have been unable to land a job. This is typical of the times, but even more so because my specific background is unheard of outside of university academia. Basically, nobody has a clue as to do with a professional background like me. Furthermore, the few positions in the Games Industry here require a substantial amount of credited work to even be considered for a job and are in the programming department (hence my first problem).

All hope is not lost however, and I still have my analytical sense and competence to handle gathered data and present it to the decision makers; hence my newfound interest in BI.

All my research surrounding BI seems to put me in a bind however. Typical BI candidates either come from a financial or statistical educational background with a little to moderate experience in coding. The coding language varies, but it seems SQL (Microsoft) and SAP are the most preferred.

What I have at most in either of those departments are my volunteer work at a local student bar as the accountant (bookkeeping) and my flair for data analysis ("data" here being everything related to the area being examined. Numbers, statements, opinions, the like).

As a last thing of note, paid/full time (re)-education is not an option. I'm currently receiving unemployment benefits and am quite dependent on it to sustain my livelihood. I have already used my "freebie" with my current education.

So my question is this: Where do I start and how do I train myself to better understand BI and the tools involved, so I can confidently say "Yes, I can do BI" given the above premise?

Radiation

If it seems like a thing you are interested in, I would try and get a meeting with someone in that field. Cold call and sort of hint at your situation. Don't go into huge details, just let them know you've been looking into the field. People like talking about themselves and what they do. They can give you insight into skills you may want to refine or things you want to watch out for. They may even know some entry level gig, or someone to talk to for an entry level.

Other than that, I don't really have anything else. I'll ask my brother, who is I think nearer to this field than I am.

Deebaser

This is actually my field.

Learn SQL and try to get a job as a "Data Analyst" or "Reporting Analyst". Without knowing SQL or being a subject matter expert, you really don't bring much to the table for a "BI" role

See if there is something on Code Academy, but your best bet is to download the (Free) Microsoft AdventureWorks database and find a tutorial focused around that.
In addition to SQL, make sure you really know Excel and Access.*


*For an entry level role, you'll likely be expected to know at least the following

excel: vlookups, applying/using filters, pivot tables, basic chart creation
Access: Left joins, creating a query, creating/using an ODBC connection.

Please feel free to batsignal me if you have more specific questions.

Yogo

Thanks, Deebaser. I've begun looking up SQL tutorials for the basics, and since you mention it, I better brush up on my excel skills. Haven't worked with Access before, but I've heard the name.

Once I've done those, I'll be sure to contact you to see what else I can learn

Aphostile

This is also my field. I'm a data scientist here in Silicon Valley.

Beyond learning SQL and HQL, I'd suggest attempting to pick up a common BI tool and begin creating analysis and reports on commonly available datasets. Some good ones to look into are MicroStrategy, Tableau or BusinessObjects.

Your research background should already prepare you for the primary goal of a BI person, which is to take data and make it tell a concise story.

minirhyder

Some tools you can learn for free: SQL, PowerPivot and PowerView (free MS Excel addons that deal with big data), Qlikview (data visualization software). All three of those use different languages. SQL is SQL, PowerPivot uses DAX and MDX, Qlikview uses its own proprietary language.

Read some books about data modeling and database structures. Those are pretty big in BI.

schuss

Another BI guy here. Learning SQL as well as having a strong understanding of how data works and how it interacts with relational and non-relational databases is key. Most of my time is spent hunting for needles in haystacks (or more appropriately, a group of similar needles in 10,000 haystacks). As others have said, pick up a BI tool (once you can use one, it's just a matter of mastering how each tool does the core stuff), I think you may be able to get powerpivot on excel for free, which should be a good (if slightly horrendous for non-star schemas) intro to the world when mixed with the new MS reporting tools.
Beyond pure tool knowledge, the biggest things you can do to give yourself a leg up are to have expertise in statistical theories and methods as well as a generally curious and driven personality. Stats will get you to solid action points sooner, while the curiosity will uncover the right things in the data.

Also, all data is dirty. ALL OF IT.

useless4

The ideal mix is mathematics, programming, visualization and subject matter expert.
If you're amazing in one or more and can build up the others you will be ok.

I for example know my data points inside outside and upside down because I am the subject matter expert for the business systems where I work. It comes naturally to me to look at 500k lines of transaction data and point out the bad one in under a minute. What I can't do is math or programming... I was kind'a led down this path because I abuse Excel with the best of them and I have a flair for storytelling/visualization.

I don't compete with the data scientist, but I play in the same pool. I do feel like I am building up my programming, my math will always be suspect, but I know I am in the game because I understand that data better then anyone else and I know how to verify/suspect the output of the dat scientist. I also come up with excellent system breaking use cases and events, which while driving everyone crazy helps move the BI to where it needs to be.

Yogo

Thanks, guys. This has really been helpful.

My personal hope is that I can combine my understanding of human behavior/storytelling with BI, so as to present the data in a manner which takes the social and human aspect into consideration. Often times I have heard or experienced a proposed change to a given system based on the acquired data without looking at the context of why said data came into existence. The soft element is often missing.

Deebaser

I don't know if many jobs purely in the "soft area" exist. Usually that role is handled by mid-level people with industry experience.

schuss

Soft element is tough, as you're often picking from millions of rows of data, so you can't analyze on a deeper level except for proof of concept work. What I would look into is text mining and unstructured data work, as that's the next hot area once people are done with moving their major metrics to proper sets of datamarts. There's a ton of useful information that's only put down in text, which drives us data people nuts.

useless4

I personally think auto weeding out resumes with greater accuracy will be the text analytics money maker. But that can be applied internally too... here are traits of successful people in our org (training classes taken, performance review key text, positions they held, position order they followed) and using that to identify similar candidates.

schuss

useless4 wrote: »

I personally think auto weeding out resumes with greater accuracy will be the text analytics money maker. But that can be applied internally too... here are traits of successful people in our org (training classes taken, performance review key text, positions they held, position order they followed) and using that to identify similar candidates.

Eh, there's a lot of other applications in medical and other fields. I work in insurance, and we're starting to use it to pull data out of the notes the adjuster leaves on the claim, or on email correspondence chains. Not to mention speeding up processing time of qualitative data from things like surveys.

credeiki

Not sure if it's legit to piggy-back on someone else's thread, but for those of you saying to learn SQL--how does one demonstrate that knowledge if it isn't used in one's workplace? I can list SQL on my resume (I've made a couple of databases for the purposes of learning, and pulled data from them, done basic operations, can figure out the rest presumably)--but that doesn't really show I can use it. Do most places have technical interviews for business intelligence where you can show your stuff? It seems like you still need to convince people via your resume first though, right?

(I'm also trying to transition into data science/business intelligence. I talked to one person in data science at a video games company who suggested that I apply to business intelligence rather than data science. I have a more quantitative/coding background than @Yogo --BA physics, PhD biophysics expected by end of calendar year--but no experience in the field. And I don't necessarily want to do big data for big pharma, although that's probably the only place where my domain knowledge is relevant.)

Rend

credeiki wrote: »

Not sure if it's legit to piggy-back on someone else's thread, but for those of you saying to learn SQL--how does one demonstrate that knowledge if it isn't used in one's workplace? I can list SQL on my resume (I've made a couple of databases for the purposes of learning, and pulled data from them, done basic operations, can figure out the rest presumably)--but that doesn't really show I can use it. Do most places have technical interviews for business intelligence where you can show your stuff? It seems like you still need to convince people via your resume first though, right?

(I'm also trying to transition into data science/business intelligence. I talked to one person in data science at a video games company who suggested that I apply to business intelligence rather than data science. I have a more quantitative/coding background than Yogo --BA physics, PhD biophysics expected by end of calendar year--but no experience in the field. And I don't necessarily want to do big data for big pharma, although that's probably the only place where my domain knowledge is relevant.)

If you want to prove something you haven't done professionally on a technical resume, do a side project involving it and then cite it on your resume.

Then they'll ask you about it in the phone screen / in person interviews, and you can really flex your knowledge of whatever you're talking about.

schuss

credeiki wrote: »

Not sure if it's legit to piggy-back on someone else's thread, but for those of you saying to learn SQL--how does one demonstrate that knowledge if it isn't used in one's workplace? I can list SQL on my resume (I've made a couple of databases for the purposes of learning, and pulled data from them, done basic operations, can figure out the rest presumably)--but that doesn't really show I can use it. Do most places have technical interviews for business intelligence where you can show your stuff? It seems like you still need to convince people via your resume first though, right?

(I'm also trying to transition into data science/business intelligence. I talked to one person in data science at a video games company who suggested that I apply to business intelligence rather than data science. I have a more quantitative/coding background than @Yogo --BA physics, PhD biophysics expected by end of calendar year--but no experience in the field. And I don't necessarily want to do big data for big pharma, although that's probably the only place where my domain knowledge is relevant.)

Depends on the role. If you're up for a programming role or expected to develop your own ETL (Extract - Transform - Load), there most likely will be some level of technical interview on it. Other roles - probably just some simple questions on it. Knowing how data structures work is far more important though, as you generally use things like informatica, SAS etc. to pull your data, which are mostly GUI-fied (if a bit odd at times). It's a field that's still sorting itself out, so there's a lot of latitude in what the role will be depending on where you land on the data analysis/statistics/presentation pyramid, as those are the 3 major facets of BI (IME).

FieryBalrog

Entry-level SQL which you'll use in a BI role is fairly easy to start learning. SQL itself has a lot of depth and complexity to it, but in a BI role you will be doing read-only access to data warehouses and OLAP cubes, which greatly limits the scope of what you need to learn. The most important thing you need to learn is how relational models (and data warehouses) work. That's actually the harder part, and something you'll probably keep learning as you do the job. Meanwhile you can learn some basic querying skills for SQL- join syntax, in-line views, analytic functions, etc.. These days I work with Java & back-end SQL, but I started in reporting doing work fairly similar to a BI, with more emphasis on writing queries and report files, and less emphasis on analysis. Learning this type of SQL is not nearly as complicated as full-on programming.

You could benefit from downloading some sample data sets and a development copy of the SQL program from the vendor of your choice- I work primarily with Oracle, and I know they have free development tools. MySQL/PostgreSQL are free and open source, and I would think Microsoft also has some free tools. (I've used SQL Server at work but we had it licensed).