Working with Steve on an inventory upload for parent attributes. We ended up talking about progress and progression. We went through tons of other topics as well. Here is a small list from my notes: Funds, budgeting, turning some corners, CPA's, financial, oversight services, different audiences and levels. Good meeting.

Eric popped in and gave us a report on his sales tax aggregation project. We covered: ETL (extract, transform, load), data warehousing, mirroring data, running aggregates against the mirrored data vs running it against live transactional data.

Eventually, everything comes down to a date or date/time (objects and data over time). We need to play to that level. We briefly talked about a 3D calendar and tracking things based on that 3D calendar (years, quarters, months, weeks, days, hours, minutes, seconds). Adilas allows you to virtually mess with time (when did things happen). Playing in real-time, that's the goal. We keep building and breaking it, we then build it and break it again and again. That is our process.

Steve was chiming in and explaining flex bubbles, flex pods, and allowing things to flex and then come back together. The discussion went towards inventory and what we know about it (it's state or status). Inventory is usually either safe and physical (on hand), on order or in transit, or sold and has a cost of good sold. Querying and flagging data (phases and sub locations) and trying to do it in real-time! Real-time aggregates plus real-time transactions (mixing them). That could create some real-time virtual smoke and mirrors.

Going to the next level - that becomes a possible value add-on layer and getting other serves (or services) involved to handle the BI (business intelligence). The whole thing deals with flags and reporting on states, status, phases, and sub locations. The goal is keeping track of changes and tracking what days get changed (virtual queue) as a grouped modified date.

As we move up the chain, we will make it faster and faster. We may also need a manual flush or way to force a recalc per day or per location. As a side note, we may even want to add a compare page between transactional data and stored aggregates.

We also talked about being able to turn things on/off as we are testing. Running experiments as needed. What about API sockets and alternate routes (we have to catch all possible routes). We talked about database triggers but sometimes that logic gets hidden.

Path - good operations (ops) lead to good accounting (acct) and that leads to good business intelligence (BI). If you keep looking at things (going forward), you get into trending and predictive analytics. Pretty cool!

Date it and flag it! Date it and flag it! - This helps us track the story of what is happening and going on. Along with that, what is the next level? Keep watching and pushing, find the edges, and try to bring things together like a zipper being zipped up - one cog at a time.