Intro
I enjoy coffee. Like, in the 6+ cups a day range of enjoyment. I’m healthy, active, in great shape so I’m comfortable with the excess. Besides, there’s an argument it drives productivity. Not exactly a strong argument, but an argument no less.
Speaking of active, more than coffee I love backpacking! I’m gonna put together some more thorough posts of the awesome places I’ve been, but for now here’s Bryce Canyon in Utah.
I’ve been involved in some form of tech my entire adult life and the last 3 years my business partner and I have been running a software dev shop focusing on custom business applications. The problem is, custom software is insanely expensive and takes a little longer than forever to complete. Along the way, we’ve built our own tools to manage and automate much of the work we do. After thinking about what we’ve built for ourselves we think we can build literally anything and so I’m interested to talk with everyone about our ideas and hear yours as well!
Speaking of what we’ve built, let me give you the quick summation.
API to SQL Database Sync
This is probably the coolest thing we’ve ever built and is definitely the most complicated. In short, we write software that calls various API endpoints to retrieve and update records which we then write to a local database. In time, we have built libraries for the primary APIs we work with to write to a local database and we realized we could make this a product.
We built out a proof of concept using the Zoho CRM API. If we logged into Zoho from our web browser, and created or updated a record, our local SQL database would be updated to reflect that change within seconds. Then, if we inserted or updated records in the database, these would appear online. Pretty effin’ cool if you ask me.
The problem here is that while our proof of concept worked it was nowhere near production ready for a customer-facing environment. And by our estimate, it would take a lot longer than forever to actually finish writing this properly.
This project sits in the background waiting for the planets to align so we can make this real.
Automated Cloud Deployments
We write web based software so we need servers. Lots and lots of servers. Our go to provider is DigitalOcean. Affordable cloud servers with a great API for automation.
While DevOps is the latest buzzword and there are countless options for cloud deployments, we’re kinda picky and like to do things exactly the way we want to. So, we built our own cloud deployment platform.
We basically write a shell script that is automatically executed at the time of deployment. Our software manages those scripts and the revision history over time so we can quickly revert to an early version if necessary.
This product is pretty mature and we use it daily so this is something we could reasonably release as a SaaS product.
Google AdWords Campaign Management
While we primarily do software development, we do general business consulting as well and we’ve run across many clients turning to Google AdWords to sell their products and find new customers. More than a decade ago I ran an ecommerce business and poured thousands of dollars into AdWords and learned everything about it.
With my knowledge and a little code, we built our own campaign management software.
This isn’t even scratching the surface here, but this not only helps us build the campaigns, it provides a data model around which we build our algorithms to maximize performance as we analyze the data.
The general idea of our platform is like this. Let’s say we have this keyword
extra large blue acme used widgets
We break this down into categories:
- size (extra large)
- color (blue)
- brand (acme)
- condition (used)
- product (widgets)
Within AdWords itself, the most granular level we can get is the keyword, but with our platform we can take a look down to the category level and the specific terms within them.
We’ve got a couple clients and this area is now my primary focus. I spend a little too much time writing SQL queries and staring at Excel. Since we’re managing multiple clients and we want to reuse these queries, we started to think about how to manage everything. And then we sat down and built it.
SQL Studio
We’ve all heard of SQL Server Management Studio, the tool used to connect to Microsoft SQL Server databases. There are also things like MySQL Workbench and phpMyAdmin, among many others. Much like cloud deployments, we’d rather have exactly what we want so built our own.
In the simplest sense, I write a SQL query and save it in a text file. Our software sees that file and adds the query to our database. If I later change that file, we see that and then create a version in our database.
In the query itself, we allow variables and named parameters, which are basically ways to allow the user to specify values at execution. In this image you can see we have the date fields and what we’ve named $varchar_group by. As the name implies, we recognize specific data types and handle them appropriately.
Once the query is executed, it is displayed in a clean searchable and sortable HTML table. (Sorry, this was real data so I had to blur it here.) We also store every execution and the result in our database, so we have a complete history of all revisions of the query and every single execution.
What’s Next
Basically, the AdWords related work has been growing so now we’re playing around with data science and looking at statistical modeling and the r programming language. I am deeply interested in blockchain and am excited by the work SteemIt has done so far.
As you can sort of tell from this cursory recounting of anecdotes we can build any kind of software that we want. These are a few of the things we’ve built and I’ll be sure to share more ideas and things we’re working on. I’d love to get feedback on the tools I’ve talked about here as well as your thoughts on what you think is interesting and excited in tech.
Looking forward to connecting with the community!