8 years inside proxy infrastructure. Started in support, ended up building residential proxy products from scratch. I know this industry from the network layer to the pricing model.
What I've actually built
Everything below I built while working at a proxy company for 8 years. Most of it I figured out from scratch, without a team or a playbook.
Buying IP blocks from LIRs and brokers, setting up ASNs, finding datacenters to announce them, configuring routing, keeping ISP classification intact across GeoIP databases. Learned by doing.
Evaluating hosting partners for IP announcement, understanding what BGP vs static routing means for classification, working with GeoFeed to correct database records. Done across multiple setups.
Built the full compliance framework: AUP, DMCA handling, abuse intake and escalation, upstream communication. Also designed the throttling system controlling concurrent sessions and RPS per client, so one customer couldn't affect neighbors on the same server.
Designed the FUP from scratch: traffic segmentation, thresholds, trigger logic, customer messaging. Then rebuilt the pricing model around it so the cost-per-GB economics worked at scale.
Finding proxy suppliers, running calls, negotiating GB-based pricing, technical due diligence before scaling. Built the full lifecycle: discovery, testing, production, rotation, exit.
Wrote specs for internal tooling: usage and traffic reports, IP occupancy tracking per client, network and provider accounting in the CRM. Translated operational pain into requirements engineering could ship from.
Researched and designed residential proxy infrastructure from zero. Studied public patents from established market leaders to understand how the underlying architecture worked. Covered SDK supply, traffic routing, load balancer logic, throttling, compliance, and the full pricing model.
Researched scraper products from scratch: how JS rendering works, antibot handling, per-domain RPS constraints, compliance limits. Wrote the product spec, tier structure, and pricing before any code was written.
Expertise
I can talk to a network engineer about BGP and to a CFO about unit economics. Not many people in this space do both.
Proxy infrastructure
IP routing & geolocation
Supplier & procurement
Product & commercial
Skills & background
Started in support, moved through procurement into full product ownership. Skill depth below reflects that progression.
Product & strategy
Infrastructure & technical
Procurement & business
Compliance & risk
AI & agentic coding
Soft skills
Conferences & industry events
Case studies
Each of these started with a gap — no system, no process, no product. Everything below was built from scratch and handed off to engineering.
Cost per GB reduced by 3-4x
Problem
Infrastructure costs were not tracked systematically. Traffic overuse was caught manually by sysadmins after the billing cycle — no FUP, no throttling, no automated controls. Cost per GB was high and unpredictable.
What I did
Result
Cost per GB reduced. Infrastructure became predictable and controllable. Supplier pipeline automated — discovery, qualification, and benchmarking no longer required manual effort.
FUP and throttling system built from zero
Problem
No Fair Usage Policy existed. High-traffic clients consumed bandwidth without limits, affecting neighbors on shared servers. Sysadmins caught abuse manually and applied blunt speed caps — no gradual response, no client notification, no automation.
What I did
Result
Manual intervention by sysadmins eliminated for standard cases. Infrastructure load became predictable. 95% of clients operate without hitting FUP. Neighbor isolation on shared servers improved significantly.
Scraper API product vision — full spec before any code
Problem
Competitor scraping products existed but nobody on the team had a clear picture of how they worked internally, who the real customers were, what features were must-have vs nice-to-have, or how to price it sustainably. No spec, no plan, no research baseline.
What I did
Result
Complete product vision delivered to CTO — architecture, pricing, GTM, roadmap, and 8 side product specs. Engineering had everything needed to start Phase 1 without a single clarification meeting.
Residential proxy architecture designed from zero
Problem
No residential proxy product existed. The team wanted to build one but had no understanding of how it worked technically, where IPs come from, how SDK supply works, what the unit economics look like, or how to price it.
What I did
Result
Full architecture spec delivered solo: SDK supply model, load balancer design, geo-routing logic, compliance framework, pricing model. CTO received a complete technical blueprint to begin implementation.
AI-powered supplier discovery tool built with Claude Code
Problem
Finding new IP suppliers and datacenter partners was done manually: search engines, blogs, forum posts, cold outreach. Time-consuming, inconsistent, hard to scale. No structured pipeline for discovery, qualification, or benchmarking.
What I did
Result
Manual supplier research eliminated. Pipeline became systematic and scalable. New suppliers identified and qualified automatically. Tool is actively used for ongoing procurement and conference preparation.
How I work
The first years I understood the product from the customer side: what breaks, what confuses people, what they need. That lens stuck, even when the job became infrastructure and procurement.
When I build something I start from the problem, not the solution. For a throttling system or a pricing model, I want to understand the economics and failure modes before writing a spec.
Before designing a product or signing a supplier deal, I check whether the infrastructure actually supports it. Routing, classification, GeoIP, cost per GB. Most product problems trace back to assumptions that weren't tested here.
A supplier that works for one product tier often doesn't work for another. I test against the actual quality bar, not a generic benchmark, and document gaps before scaling any traffic.
Pricing and cost structure need to be figured out before features. I've seen products that couldn't be priced sustainably because the infrastructure assumptions were wrong from day one.
Coming from support and ops, I know what happens when a requirement is vague. The specs I write cover edge cases and failure modes, not just the happy path.
Looking for
Looking for a proxy, data infrastructure, or IP services company where someone with this background can step in at a senior or leadership level and move things forward.
I've built most of this before. Fewer mistakes, faster ramp-up, less time explaining why something is harder than it looks.
Positions I'm targeting