WISP · FISP · Rural broadband — Central Texas
I’ve built and run ISPs since 1994.
I help wireless and fiber ISPs fix unreliable networks and plan for growth. I’ve grown my own ISP past 10,000 subscribers and expanded a four-tower network to thirty towers serving thousands, while maintaining 99.99% uptime.
If your network is a mess right now, I’ve almost certainly fixed worse.
- 1994
- Building ISPs since
- 10,000+
- Subscribers grown
- 4 → 30
- Towers, one network
- 99.99%
- Sustained uptime
Region’s first provider, out of a BBS
My own ISP, from zero — as the owner
100 customers to several thousand
Customer-weighted tower availability
Track record
I’ve built the networks and run the business.
I started in 1994, helping a friend turn a bulletin board system into our region’s first internet service provider.
Then I built my own ISP, growing it from zero to more than 10,000 subscribers. As the owner, I was responsible for the network, the payroll, the truck rolls, and the 2 a.m. phone calls. I know what these problems cost because I’ve carried those costs myself.
Most recently, I took over a four-tower network with about 100 customers and grew it to 30 towers serving several thousand subscribers. I rebuilt the network with semi-mesh fiber, wireless backhauls, and a dual-homed BGP core. Customer-weighted tower availability exceeds 99.99%, with 100% core routing availability over the past year.
-
01 Cut the truck rolls
Reverse-engineered the radios and built access-layer diagnostic tooling — cable fault and SNR, Ethernet negotiation, DHCP/ARP correlation, PCAP session visibility, latency and jitter. Techs stopped driving out to diagnose what could be read from the office.
-
02 Built the NOC, not just the network
Stood up a NOC and SOC with standardized incident response on Zabbix and Elastic, with custom triggers for BGP events, backhaul misalignment, and AP crashes. Reliability stopped being a feeling and became a trend line you can plan capacity against.
-
03 Beat the public resolvers
Deployed distributed DNS caching with DNSSEC validation at the fiber transport sites. It measurably outperforms Cloudflare and Google for our subscribers, which is the cheapest perceived-speed win in this business.
-
04 Owned the whole build lifecycle
Tower and fiber, start to finish — siting and acquisition, structural and steel, OSP fiber for FTTH, DC plant and inverter power systems, through to turn-up and ongoing maintenance. Not coordinated. Owned.
The core offer
Any network. Any condition.
Undocumented. Inherited from someone who left. Three generations of gear from five vendors, half of it configured by hand at midnight. Flat layer-2 that falls over when it rains. I’ve walked into all of it. The condition of your network is not a blocker — it’s just the starting point.
I show you the path to 99.99% uptime.
What I fix
I know this industry at every layer of the stack.
Most consultants know one slice — the radios, or the routing, or the billing. The problems that actually take an ISP down live in the seams between those slices. I work the whole stack, which is why I can find them.
RF engineering at scale
- 80 GHz, 60 GHz, 18 GHz, 11 GHz licensed and unlicensed backhaul
- 5 GHz and CBRS access layer
- Tarana, Ubiquiti, Cambium, Aviat, Mimosa
- Path and link budget design, interference hunting, spectrum planning
- Backhaul misalignment detection and capacity forecasting
Core, routing and transport
- Multi-homed BGP transit and peering design
- Dual-homed core, semi-mesh fiber and wireless backhaul
- Carrier-grade redundant topology design
- DNS and DNSSEC caching infrastructure
- Traffic engineering, QoE, Preseem-class shaping
Towers and steel
- Site acquisition, siting and lease negotiation
- Tower construction, structural assessment and maintenance
- Climbing and rigging, mount design, loading analysis
- Turn-up, commissioning and ongoing inspection
Fiber and FTTH
- OSP fiber construction and route design
- FTTH build planning and GPON architecture
- Splicing, OTDR testing, documentation
- Fiber transport to tower and data center sites
Power and plant
- DC plant design and build
- Inverter and rectifier systems, battery sizing
- Generator and backup power, runtime planning
- Grounding, bonding and lightning protection
Software, automation and the NOC
- Monitoring and observability — Zabbix, Elastic Stack, custom triggers
- NOC and SOC standup, incident response process
- Billing and OSS integration, provisioning automation
- Custom tooling — CPE diagnostics, network automation, internal APIs
- Virtualization and storage, backups, retention for compliance
How it works
Here’s what you get.
You should know what you’re paying for, what I’ll deliver, and who owns the work. I start with a written assessment, agree on scope and price before implementation, and hand over configurations and documentation as the work progresses. No retainer required.
- 01
Written assessment
You get a clear written assessment of what’s wrong, why it’s happening, and what it will take to fix it, and what it will cost — without needing an engineer to translate it first.
- 02
Fixed scope, fixed price
You know what I’m doing, what I’m delivering, and what it costs before work starts. If something changes the scope, we agree on the change before the work moves forward.
- 03
You own everything
Every configuration, credential, diagram, and document is yours. I hand them over as we go, so you have what you need to run your network with or without me.
- 04
No retainer required
You hire me to solve a specific problem. When it’s solved, the project is complete. Ongoing support is available if you want it, with no obligation to keep me on retainer.
- 05
I’ll tell you if you don’t need me
Sometimes the answer is a firmware update, a configuration change, or a call to your existing vendor. If that’s all you need, I’ll say so. I won’t turn a simple fix into a bigger project.
Get in touch
Tell me what’s broken.
A few sentences is enough to start — roughly how many subscribers, roughly how many towers, and what’s hurting. If it’s an outage happening right now, say so in the first line and I’ll move it to the front of the queue.
Or email directly
contact@wispstack.comGot it.
I read these myself and reply personally, usually within one business day. If this is an active outage and you don’t hear back soon enough, email me directly.