Hi everyone, it's been a while so shoot me if I'm taking a liberty here... but I don't know where else to turn.

"We" (as in my me and my employer) are after "an IT Person" - but haven't the first clue how to best frame the job advert.

We have (external) IT, who are more than up to the task of day-to-day systems management, but internally we feel a need for someone in-our-(direct)-employ, not only to better liase with said (external) provider, but to also (mostly?) strategise and develop ideas and workflows with our own (architectural) teams.

With my own work moving ever more away from the job-floor (that is architecture) and into quality management and compliance, I just haven't the capacity to commit to learning all the tricks I (we) need. Specifically; we have a collection of business-line software we wish to better integrate (all come with an API) and I have (potentially blue-sky) aspirations to pool them into a data warehouse with a mind to start actively employing data analysis tools like Power BI, automate report production, etc etc.

The thing is, it is (relatively) easy for me to list the things I think we need into a advert sentence of the like:

"Candidates will have proven and demonstratable working knowledge of working with Microsoft Server environments (specifically Active Directory & Group Polic managment essential), will be comfortable working with SQL & PHP, have experience working with multiple platform APIs and REST protocols and a keen interest working with big data and ideally is familiar with Power BI."

BUT

Aside from potentially short-changing ourselves with a such-a-finite list (I've already missed out ODBC, Python & C#) and not even knowing what to actually ask for (is it a 'developer' we're after?) where we would post said advert, what financial expectations we should have/make, and how we then actually interview respondents, is to say the least, beyond us.

We don't even know if we are an attractive proposition!? We're not even sure if this would be full or part time!? It would be so much easier if we chanced upon an architect-with-leanings-toward-IT (and have nods in our current adverts for architectural positions to that affect) - but they're as rare as the proverbial rock horses'...

So after another of my lengthy rambling thread-starters, the quick-to-the-point question:

Q. How might one go about hiring an 'IT person' (for an architectural practice)?





PS: Some of you might have spotted me posting an abridged version of this on Facebook some time ago (which went nowhere), my Twitter's to-tied to "me at work" to use, and I daren't post this on LinkedIn as I'll get swamped by cold-caller-leads. I did sign up to Gun.io but they got the boot when they didn't even call for the first appraisal - as for Hired, well we're trying to get away from the agents as it is.

Sounds like you're looking for somebody with devops skills but not in a traditional devops role.
IT people who do adetc usually will do Exchange or SQL but not a lot of programming languages as well.
Sounds like you'd be better off with more of a developer who's got some infrastructure skills.

Cheers Simon, sounds a lot like what I've been reading and getting the feeling from elsewhere.

To be honest, having had a really productive day, and the nights are getting longer I'm thinking it's time to knuckle down again at my home studies. (continued in next post)


That said, it dawned on me today that I might already know someone. An industry friend has enlisted the services of 'another' to develop (?!) an online platform that does a lot of similar things to those I should like to do (albeit with different data). I'm going to try and get hold of that 'other' for a chat. I don't imagine for a second "a job" will be an attractive proposition, but a commission just might?

Thing is, what do these kind of 'developments' cost? Piece of string question of course. But by way of example, whilst accepting "I could find the answer on countless forums for free if I JFGI'd better", what if:

Brief = Write a (PowerShell?) <script/routine> that runs every <day/week/month>, at <??:??>, collects <x,y&z> & <returns/imports/exports> to <?> a <.csv/.xls/.mdb/.?>.

Cost= £?

Even writing that down has made me think perhaps I'm making too much of this?

By my estimation, I have five key 'platforms' (if you will) presently before me. Access, SQL, Flux, Dynamo & PowerBI.

Access (only) because it's what I'm presently using and mildly comfortable with. I get all the concepts, and it ties nicely with one of our core information sources (Revit), is free in that it comes with Office, outputs/links to familiar formats (.xls, .doc) and at the simplest level can readily align to our stationary.

SQL because I understand its where/how my data, presently gestating in Access, should really 'be'. But whilst I understand SQL as a "I want to ask this" language, I've never really grasped where/how, as in as what, in what, the tables of 'pure' SQL live. Does that make sense? In Access you have tables, queries, forms and reports all wrapped up in a container database file (.mdb or .accdb) but "in" SQL? On an "SQL Server"? What's on "that" though? Confused doesn't even come close.

Flux, for those that don't know, is a pet project from Google, it's a free (at the moment ) "data-linker" kinda thing - almost like a relationship table all on it's own, enabling the porting of keys from one application to another (at least those presently supported, which thankfully feature, given its the target market, AEC software) to faciliate near-live data exchange. Think MSPaint tools drive an Excel PivotChart or Sketchup model.

Dynamo, for those that don't know, is a pet project from Autodesk, it's a free (at the moment ) visual-programming application, that shares similarities with Grasshopper, but with more of a leaning toward Autodesk platforms. It is immensly powerful, (moreso if you've C# and/or Python up your sleeve) and I really need to sink my teeth in.

Both Flux & Dynamo flirt in that murky space between clandestine commercial competition and community orientated knowledge sharing and positive growth. Thankfully online communities are growing exponentially around them both, so it's basically all on me.

PowerBI? Well to be honest this I'm least sure on, at least in terms of being a dead certainty. I've been long seduced by Tableau, but it's priced way off Microsoft's mark, and to be fair, I need a low-impact level of entry for this (rather pet) project of mine! I've seen some really tasty data visualiser scripts online that can be tested and had for free to use in sites (thankfully we've got a guy at work who likes that webby-side of things) - but in order to use those the project would have to quickly move into a browser. But then I haven't really got round to thinking about the UI end of things - my mind stuck on a ghetto wireframe of an interface made with Access forms! The dashboards (of PowerBI) really appeal, but I am going to need an input-frontend of some kind. It's probably this part that's the real sticking point in framing this whole thread's topic.



I have prattled here.

Developers usually are up for contract or project work. What it costs is another matter.
The more spec and info you give them the less it usually costs (give them tasks and they're quick, make them design it from a rough idea and you're talking thousands).

So what it sounds like you need is a database (SQL, MySQL, noSQL etc) to store all your data (like you do with access).
You'll need some sort of Front End to access this data visually.
Then you'll need some queries created to pull the data you want whenever you want it.

If you've got the database already in Access then you can get that converted to SQL, this could be accessed from Access or a web front end.

BI is best to replicate the database out to another BI database then query that copy

Pray tell why? Because of the "call limit" PowerBI licenses exert on use/users? Performance? Something else? All of the above?


As for design? Well thankfully that's where I like to think I might shine.

Main reasons are performance & security.
If you've got complex queries running on a database it'll slow it down and the people accessing the database for BI aren't usually SQL admins so you don't want them running queries on the live database in case they do something wrong and trash it.
Most DB's I work with have thousands of users at any time so a rogue badly written BI query sucking all the servers power is a disaster