Turning repetitive infrastructure work into software.
Owned the roadmap for an internal GIS/CAD network design tool used by 200+ engineers, and turned repeated operational bottlenecks into automation adopted company wide.
200+ engineers designed networks in a GIS/CAD tool, but recalculation, BOM pulls and error checks were all done by hand around it.
Interviews with 30+ designers kept naming the same two time sinks: recalculating signal levels after a change, and assembling bills of materials.
A three click Python automation for tracing, recalculation and verification, and an Auto BOM that records equipment changes as they happen. One team first, then three regions.
Recalculation 90% faster, manual errors down 25%, design turnaround 40% faster, Auto BOM used company wide, about $700K saved per regional operation.
What I owned
- Product owner for Spectrum's internal GIS and CAD network design tool used by 200+ engineers, translating feedback from 30+ users into prioritized improvements and aligning Design, Construction, Operations, and Engineering teams, cutting design turnaround time 40%.
- Identified manual downstream signal recalculation as a key design bottleneck and defined a 3 click automation for tracing, recalculation, and verification: recalculation time down 90%, manual errors down 25%, about 2 engineering hours saved per project across 3+ regions.
- Identified manual BOM generation and equipment validation as recurring pain points through designer interviews and launched Auto BOM, which tracks equipment changes automatically: report generation time down 50%, better network upgrade budgeting, about $700K in savings per regional operation.
- Owned the prioritization model for a 30 person cross functional unit, ranking requests by volume, urgency, and dependency.
- Led a 20 person audit of 1,000+ optical nodes for DOCSIS 4.0 readiness.
What had to be true
- Any automation had to slot into an existing, already adopted GIS/CAD workflow. Engineers would not switch tools, so the tool had to come to them.
- Errors in signal recalculation or BOM data have real field consequences, so automation had to be at least as accurate as careful manual work, not just faster.
- Prioritization had to serve a 30 person cross functional unit with competing, legitimate priorities.
Finding the bottleneck through interviews
Rather than assuming what to automate, I interviewed 30+ designers about where their time actually went. Manual downstream signal recalculation and BOM and equipment error checking came up repeatedly, independent of region or seniority.
Designing a three click automation, not a new system
Because designers would not adopt a separate tool, the automation was built as a lightweight layer that traced, recalculated, and verified signal levels in three clicks inside the existing workflow, cutting recalculation time significantly and manual errors by 25%.
The same pattern, trace, compute, verify, was reused for Auto BOM, which recorded equipment changes automatically instead of requiring a manual pull.
Scaling adoption
Both tools were rolled out incrementally across regional teams. The recalculation automation was adopted by 3+ regional teams, and Auto BOM went company wide once report generation savings of about 50% were demonstrated in the first region.
What shipped
- A three click Python automation for downstream signal tracing, recalculation, and verification, embedded in the existing design workflow.
- Auto BOM: automatic recording of equipment changes as designers made them, replacing manual bill of materials assembly.
- A prioritization model for the 30 person cross functional unit that ranked requests by volume, urgency, and dependency instead of first come first served.
How it was built
- Python automation layered on top of the existing GIS/CAD tool's data model, so designers kept their existing workflow and tool of record.
- Change tracking hooks that captured equipment modifications at the point of design, rather than reconstructing them later from disparate records.
Trade offs I made on purpose
- Prioritized automating the two highest frequency manual tasks (signal recalculation, BOM pulling) identified directly from user interviews, over building new design features.
- Chose incremental regional rollout over a single company wide launch, using early results to build the case for wider adoption.
Takeaways
- The highest leverage product opportunities were hiding inside operational workflows nobody thought of as a product. They just looked like how the job was done.
- Automation that respects an existing workflow spreads faster than automation that asks people to change how they work.