The Last Mile: When Technology Disappears into the Work

Neil Chester

Nobody remembers the day new software gets installed. People just remember the times it gets in their way. Maybe it was the Revit model that ran slowly or wouldn’t open at all over a VPN from a hotel room. Maybe it was the spreadsheet someone had to rebuild every week using existing data. Those moments may seem small, but they add up and shape how people feel about every tool that comes after.

As Director of Technology at Apogee, I see my job as eliminating those moments. I call the goal frictionless technology, which means planning every tool around the person who will use it. It starts with understanding the work well enough to identify the right solution, what Stephen Covey (1989) described as “seeking first to understand.” Then comes the part technology teams tend to overlook. People need to understand how a tool benefits them before they are willing to change how they work, and they need sufficient training to ensure the investment is fully utilized. Without these steps, even good software can become something people work around.

I’ve also come to recognize no single platform does everything. Vendors develop solutions for the broadest market possible, which makes sense for their business but leaves a gap for everyone else. I think of that gap as the last mile, and in a firm our size, closing it falls to our IT department. It’s the difference between software that technically works and technology that actually works for the people using it. Closing that gap often means adapting existing tools, connecting disconnected processes, or removing unnecessary steps from everyday work.

Sometimes closing that gap starts with making sure technology works wherever our employees do. Revit and many of the other design tools our industry relies on still require high-performance hardware, yet our employees regularly work from hotels and job sites. Their workstations live in a data center, and the laptops they carry provide remote access. We designed this setup so employees could travel light, but it proved most valuable during unexpected disruptions. When the office lost power, everyone was able to pack up, relocate, and continue working. When a laptop died, we were able to provide a loaner, and within five minutes, that employee was back on their own desktop exactly as they had left it.

Some may ask why we don’t simply move everything to Autodesk’s cloud collaboration tools. We currently use Forma Design Collaboration when we’re serving as a consultant on another firm’s team, and we may lean on it more in the coming years. However, it solves a different problem. Cloud worksharing determines where the model lives, while VDI determines where the computing takes place, and Revit still requires a powerful machine somewhere. For us, the priority isn’t adopting a particular platform; it’s ensuring employees have reliable access to the tools they need, regardless of where they’re working.

Other times, the last mile is about eliminating unnecessary work between systems. Our accounting system had every number we needed to manage collections, but it presented them as one long list, and turning that list into a daily plan meant rebuilding it in Excel. That process cost hours every week. Instead of replacing the system, we built a small browser extension that works within the existing interface and displays only the invoices requiring a call or email that day. It records follow-up notes directly into the accounting system, eliminating the need to maintain a separate database. Bell and Orzen (2010) frame Lean IT around removing waste from processes. A spreadsheet that exists to make existing data accessible is a clear example of unnecessary work.

Increasingly, the last mile is the space between people. Our CFO, John Ramseur, recently wrote about the timeline from order to cash. I see the same timeline from the technology side, as one continuous pipeline from the first lead to finished building. Currently, that process involves numerous handoffs where information must be manually transferred from one person or department to another. Each handoff creates another opportunity for duplicated effort, lost information, or delays. Closing these gaps means looking beyond individual software applications and considering how information moves across the organization. We’re working to streamline these workflows one process at a time, starting with the report our marketing team prepares for its weekly Go/No-Go meeting.

Here’s what I like about that approach. Marketing was already experimenting with AI to prepare the report, uploading solicitation documents into a chat tool with prompts they had refined themselves. I’ll take that kind of adoption every time. Our role is to turn their experiment into an automated workflow. When new solicitations come in, AI will review them and flag anything that could prevent us from submitting a proposal, with a summary ready before Monday’s meeting. The decision remains with people. AI’s role is to make sure they walk into the meeting with the information they need to make an informed decision. The goal isn’t to introduce AI for its own sake, but to build on a process employees have already identified as valuable and remove the repetitive work surrounding it.

Of course, making technology easier to use also means being deliberate about how it’s implemented. We’re careful about how we approach AI adoption. Staff use a limited selection of approved AI tools, and only publicly available information is entered into them. This excludes more information than people might expect. Task orders under our existing contracts are distributed only to the firms holding those contracts, and some public postings include controlled drawings that require government approval to download. Therefore, we start with information that is truly public.

I also make an effort to consider a range of outside perspectives. Theo Browne keeps me honest about whether software is actually effective and user-friendly, Eli the Computer Guy encourages me to consider what happens when it fails, and Nate B. Jones brings the conversation back to practical workflows and what successful implementation looks like. Together, those perspectives reinforce a simple principle: successful technology isn’t defined by what it can do, but by how reliably and effectively it supports the people depending on it.

John Doerr (2018) makes the case in Measure What Matters for evaluating work based on its outcomes. For us, that means measuring the hours we hand back to employees. The number of AI tools we roll out is beside the point. What matters is whether we’re removing obstacles, reducing repetitive tasks, and giving employees more time to focus on the work they were hired to do.

Our engineers already understand where this is headed. Nobody in a hospital thinks about the ductwork until the air stops moving, and that’s exactly how it should be. I want technology at Apogee to function the same way, allowing a designer or a project manager to get through an entire day without having to think about technology once. When we finish the last mile, the technology disappears into the work.

References

Bell, S. C., & Orzen, M. A. (2010). Lean IT: Enabling and sustaining your lean transformation. Productivity Press.

Covey, S. R. (1989). The 7 habits of highly effective people. Free Press.

Doerr, J. (2018). Measure what matters: How Google, Bono, and the Gates Foundation rock the world with OKRs. Portfolio.

Ramseur, J. (2026, September 28). The future of A/E accounting. Apogee Consulting Group.