ISP and WISP Network Engineering Support
Small ISPs and WISPs run carrier-class networks with a handful of people. Engineer Austin, based in Georgetown, Texas, fills the senior network gap: designing and documenting the core, edge and access network, planning BGP, CGNAT and IPv6, towers, backhaul and fiber, hardening and monitoring, and standing behind your staff as the escalation point when something breaks.
Core, edge, access and outside plant
- Core and edge design. Point-of-presence and failure-domain layout, redundancy, and migration from flat switching to a routed core.
- BGP, peering and transit. Multihoming, route policy and prefix filtering, transit and peering strategy, and ASN and address requests.
- CGNAT and IPv6 planning. CGNAT sizing, logging and failover, plus an IPv6 plan and dual-stack rollout that shrinks CGNAT over time.
- Access networks. Fixed-wireless design on MikroTik, Ubiquiti and comparable platforms: sector and channel plans, interference analysis, point-to-multipoint backhaul.
- Tower and backhaul planning. Path studies, licensed and unlicensed microwave, ring and redundancy design, with structural coordination through our tower services.
- Fiber and PON design. Splitter architecture, OLT placement, fiber route and count design, and outside-plant coordination, with construction through our fiber service.
- NOC, monitoring and capacity planning. Monitoring design, alert thresholds, out-of-band management, configuration backup and traffic analysis.
- RADIUS and billing integration. Authentication, DHCP, shaping and billing tied together so provisioning and suspension need no technician.
- Security hardening and scripted cutovers. Reviews against a written standard, then staged changes with pre-flight snapshots, rescue rules and a tested rollback.
- Senior network support on retainer. Design review of staff proposals, vendor-neutral procurement advice and on-call escalation.
We also provide regulatory and grant engineering support: the network designs, diagrams, cost schedules and exhibits that filings and funded builds ask for, with professional engineer certification where a program requires it.
Audit, design, stage, cut over
Discover and audit
Read-only first. We document topology, routing, addressing, platforms and monitoring as found, then deliver numbered findings ranked by risk. The report is yours either way.
Design and roadmap
Reference architecture, addressing plan, routing policy, access and outside-plant plans, security and monitoring design, in a roadmap tied to your budget. Projects are fixed-fee; support runs on a retainer.
Lab, stage and cut over
Templates are tested before they touch production. Migrations run in maintenance windows beside the old equipment, with failover checks and a rollback script ready before the first command.
Document, train and stand by
As-builts, runbooks and operator training close each project. On retainer we take escalations, review capacity and lifecycle quarterly, and keep the architecture document current.
Designs, templates and runbooks in your name
- Architecture document with layer 1, 2 and 3 diagrams.
- IP address plan for IPv4 and IPv6.
- BGP policy document and prefix-filter set.
- CGNAT sizing memo.
- PON design with splitter maps and fiber counts.
- Path-study and link-budget reports.
- Security baseline and hardening checklist with numbered findings.
- Monitoring design with alert thresholds.
- Configuration templates, cutover scripts and rollback plans.
- Runbooks, operator training and a monthly retainer report.
How AI speeds this up for you
Capacity models, BGP policy and IPv6 address plans are drafted and cross-checked with AI assistance against your as-found data and our written standard, then reviewed by the staff member responsible before anything is staged. Design documents reach you sooner, with every assumption written down.
Carrier-class design, independently reviewable
Our principal, Gregory "Lynn" Haag II, P.E., started in 1997 on the support desk of a mid-sized ISP and worked up to system administrator; he has since consulted for ISPs across the United States and in Mexico, Panama, Greece and India, and sets the standards our team uses to build and run the firm's own core as an ISP-class network.
Carrier design here is done on engineering principles under a licensed engineering firm: capacity and failure-mode analysis before a routing policy is written, assumptions and revision history carried on every design so another qualified person can review it, and independent review of an existing design on request.
One firm covers the RF path, the tower, the fiber plant and the BGP policy, and where a tower needs a sealed structural analysis, we say so and arrange it. The reference architecture is revised after every engagement, and PON design practice, routing security and failover over LTE and Starlink are re-evaluated each year. Engineer Austin is Texas Registered Engineering Firm F‑15739; our team does the work to house standards set by our licensed Professional Engineer.
Platforms we build on
MikroTik RouterOS for routing, CGNAT and the subscriber edge; Ubiquiti UISP and airMAX for point-to-multipoint access and backhaul; WireGuard and IPsec for management and site-to-site tunnels; Starlink and LTE as backup paths. Recommendations are vendor-neutral, and configuration and documentation stay in your name.
Who this is for
Wireless ISPs with one or two network staff, regional providers adding fiber beside wireless, utilities and districts starting a broadband service, and multi-tenant properties that behave like a small ISP.
ISP engineering, answered plainly
We have one or two network people. Can you back them up without taking over?
That is the normal arrangement. Your staff keep the day-to-day work and the credentials. We review their designs before changes go in, take the escalations they cannot clear, and handle projects that need senior design time for a few weeks. Everything is documented in your systems, so your team gets stronger rather than more dependent.
Should we get our own ASN and address space, and what does multihoming involve?
If you have or plan a second upstream, yes. Your own ASN and prefixes let you announce the same addresses through either provider and fail over without renumbering customers. Multihoming means a BGP session to each upstream, a policy for what you announce and accept, prefix filtering and a tested failover. We prepare the registry requests and the policy.
How do we roll out IPv6 and shrink CGNAT without breaking customers?
Dual-stack, one segment at a time. We write an IPv6 plan with room to grow, enable it on the core and edge first, then on access segments and customer gateways in stages, watching traffic shift off CGNAT as content moves to IPv6. CGNAT stays for IPv4, sized and logged properly, and shrinks as the rollout proceeds.
Can you design a fiber build next to our wireless network and help with the grant paperwork?
Yes. We design the PON architecture, splitter layout, fiber routes and counts, and coordinate outside-plant permitting, pole attachments and tower work so fiber and wireless share one plan. For funded builds and filings we prepare the network designs, diagrams, cost schedules and exhibits the program asks for; your grant consultant or counsel handles eligibility and the application.
What does a hardening review cover, and what are the first fixes?
Management-plane exposure, firewall and filter order, BGP prefix and bogon filtering, anti-spoofing, IPv6 posture, remote access, logging and backups, checked against our written standard. The first fixes are nearly always the same: lock management to known addresses and out-of-band paths, filter what you announce and accept, and back up every configuration automatically.
How do we tie RADIUS, shaping and billing together so provisioning is automatic?
The billing platform becomes the source of truth. When an account is created, paid or suspended, it tells RADIUS, and RADIUS tells the router what address, speed and state the subscriber gets. We design that chain for your billing platform and edge equipment, test it on a lab subscriber, and document it so a new hire can follow it.
What does on-call escalation look like, and how fast do you respond?
Retainer clients get a direct escalation path to senior network staff who already know the network. During business hours we respond within two hours. After-hours and emergency response are by arrangement, with a four-hour callback target and a one-hour target for declared emergencies.
Start with a written assessment.
Tell us what you have and what is not working. Someone from our team, not a salesperson, replies within one business day by email or phone, and brings in our engineer when your request needs one. The assessment report is yours to keep whether or not you buy anything.