Bluntman to the rescue 
You said thats your first “professional” project, however I am uncandied as usual, not to offend you but to help you improve 
The main question is what the requirement of the project where.
If it was photorealism to mime a photo of the finished building it´s a fail.
If it was a technical render, just to visualize how the building would change the appereance of the area and give an indicator on what it would look like it´s cool.
But as it was a contracted work and the client obviously payed it´s a win. And it´s a good thing to build a client base, especially for architectural stuff. So be sure to improve as well, word of mouth is the best advertisement you can have in that profession. In the construction buisness is a lot of favourism, so if you impress one of them, he´ll make sure others are impressed too 
Things to improve (some already mentioned) and random brabbel:
Choose your weapon wisely and know how to use it. If you know your way in Blender Internal it can achive almost photo realism, however your render does not. To me it looks like a better OpenGL render. For what I see you haven´t even used AO?
You might want to consider Luxrender, Indigo, Yafaray or Octane, where latter one is not free. Try them out and use what gives you desireable results the easiest way in shortest time.
Read up on lightning. I looked on your homepage, and lighting is a huge weakness on your side. I recommend Digital lighting and rendering it is a great book about the topic and software independend.
Your render does not match up. The background image has a whole different mood than the composite model, it seems to be a cloudy day in the afternoon, with diffuse bluish light, you choose a warm sunlight with hard long shadows, something that occurs at around 10am in summer - It´s very hand to use a sunlight simulator/sun system for archviz, like Yafaray, Octane has, there is even an addon for Blender for it. Where you can set geographical longtitude and latitude, choose the day of the year and the time of the day.
With proper lighting you can start to improve material and textures. Your model looks like painted cardboard. Be careful and niggling with the materials, you can always re-use them. It´s not like concrete is to change a lot over night 
Not FOSS but free, a good asset to archviz, Draftsight. It´s a free AutoCad clone from the makers of SolidEdge.
As archviz is often printed, you might want to calibrate your screen (for budget you could get a Pantone Hyue Pro or a Datacolor Spyder 3 Pro. I went with Datacolor, I like the hardware better and they offer on demand software upgrades.) Mine are calibrated to 90cd/m2 (i work in a cave
), whitepoint at 5800Kelvin and gamma at 2.2 and it looks rather bleak and dark at the same time.
In archviz often the clients simply want to print their stuff and hand it out, not correct it before printing. Especially if you get RAL colors for the plaster. If the printing device uses a color profile it´s usually calibrated good enough and the print will not match what the client sees on his uncalibrated screen, but the print will match the original RAL colors. They´ll be happy if you tell them if they complain the colors do not match, that its the clients fault and that the print looks exactly like the house will look if they stand in front of it on 2nd July at 8am. For such stuff I often use keydates.
Like for construction plant prototypes I choose the date of the company founding, or of the start of the prototype building.
For architecture you can choose the day of the hand-over of keys. Then you can tell the the building project organizer they can tell their customers “…and thats how your flat will look like when you hold your keys in your hands if you sign now”
Help the client sell his product, your client will help you sell yourself.
Like Michael already said… informational… sell the building. Usually archviz stuff is the advertisement of the building. Exaggerate.
I´ll stop here - hope there´s some advice you can use.