// STEVE HANOV — TECHNICAL CONSULTANT

YOUR DIAMOND PICKAXE FOR THE PROBLEMS THAT MATTER.

If you're a technical founder who's outgrown your first architecture. An investor who needs to know if the code matches the pitch deck. Or a company facing a problem that sits between domains — where the firmware engineer says "talk to the AI team" and the AI team says "talk to the product people." I work across all three. 25 years of shipping real products, $0 in VC.

[ START QUEST ] [ VIEW INVENTORY ]

◆ What I Do

I'm most useful when the problem is ambiguous, cross-disciplinary, or stuck.

My range is unusual because I've always worked across the whole stack. Sometimes it means designing AI agentic workflows. Sometimes it means helping a SaaS founder figure out what to build. Sometimes it means building a routing engine from OpenStreetMap data. If your problem doesn't fit neatly into one specialist's job description, this might be for you.

🤖

◆ AI Strategy & Agentic Workflows

Your AI investment ships to production instead of dying in proof-of-concept hell. Agentic workflows, voice agents, and LLM tools that create measurable leverage — not slide decks.

📡

◆ Embedded Systems & Protocols

Your IoT product ships on time because the radio stack actually works. Drawing on my cellular protocol background (EDGE/LTE) to design and debug what sits between your hardware and the network.

🏗️

◆ Software Architecture

The decisions you make in the first 90 days compound for years. I help you make the right ones — architecture, infrastructure, and technical strategy that scales with you instead of against you.

📈

◆ SaaS Product Strategy

What to build. How to position it. How to get it in front of the right customers. Product strategy that comes from 16 years of doing it myself — not from a framework I read in a book.

🔍

◆ Technical Due Diligence

I read the code, not just the pitch deck. Investors and acquirers get an honest technical assessment — what's solid, what's technical debt, and what would take six months to untangle if you bought it tomorrow.

🛟

◆ System Rescue & Cost Reduction

Your AWS bill dropped 40%. Your over-complicated system got simpler. Nobody had to rewrite it from scratch. Sometimes the fix is smaller than everyone thought — but someone has to know where to look.

◆ R&D Tasks

You can't vibe-code the hard problems.

When the obvious approach isn't fast enough, when the off-the-shelf library breaks at your scale, when the research paper exists but nobody has turned it into working code — that's where I come in. I read the papers, pull out the practical algorithms, and ship working solutions.

◆

◆ CRDTs from Scratch

When Yjs wasn't cutting it for a real-time collaboration problem, I implemented a Conflict-Free Replicated Data Type from the academic literature. Not wrapping a library — building the data structure from first principles so it actually worked at our scale.

◆

◆ Vector Search at Speed

Standard vector database techniques were too slow for the dataset. I pulled algorithms from cutting-edge research papers and optimized the search pipeline until it ran at interactive speeds. No off-the-shelf solution was going to cut it.

◆

◆ Speech Recognition Tuning

When speech-to-text kept mangling names and domain-specific terminology, I solved it. Sometimes the gap between "works in general" and "works for your use case" is a research problem — and I've solved that one before.

How I Work

A conversation, not a sales pitch.

YOU SEND A SIGNAL

Send me a message about what's broken or what you're trying to build. I reply within 24 hours.

WE TALK

A 30-minute call. No pitch, no pressure. You describe the problem. I'll tell you honestly if I can help — and if I can't, I'll point you to someone who can.

I DIAGNOSE

If it's a fit, I do a lightweight diagnostic — code review, architecture assessment, or strategy session. You get concrete findings before any major commitment.

YOU GET RESULTS

Concrete output: a roadmap, a prototype, an audit report, or direct implementation. No surprises. No bills for "strategy alignment sessions."

◆ Boss Battles Won

Real problems I've solved — for clients and for myself.

ALGORITHMS • MANUFACTURING

One algorithm. Zero wasted granite.

A SaaS company serving the granite countertop industry needed a way to optimize slab cutting. Drag-and-drop shapes weren't enough. I built a genetic algorithm with no-fit polygon caching — running in the browser — that nests irregular shapes into slabs with minimal waste. Then solved seam placement: the algorithm optimizes for both material efficiency AND visual beauty, working around fossils, veins, and natural stone patterns.

Polygon nesting at interactive speeds — in browser

The founder had been in the industry for years. Nobody had told him a computer could optimize this automatically. Four years of continuous improvement — the algorithm kept getting smarter.

INFRASTRUCTURE • LOGISTICS

They were paying Google Maps. Now they run their own for $40/month.

A Canadian logistics company dispatching trucks nationwide was getting squeezed by Google Maps API price hikes. They were falling back to spreadsheets. I built them a complete routing engine using OpenStreetMap data — then squeezed it onto a $40/month server with just 4GB of RAM. Wrote a thin Go wrapper that mimics the Google Maps API so their existing code didn't need to change.

$40/month replaces Google Maps API entirely

Geo data pipelines, routing graphs, waypoint optimization — all running on hardware that costs less than a dinner.

SAAS • UML

WebSequenceDiagrams

One of the web's most widely used sequence diagram tools. Converts plain text into professional UML diagrams. Built and run solo for 12 years.

$10K+ MRR on a $5 VPS

Single Go binary + SQLite. No microservices. No Kubernetes.

AI • FINTECH

eh-trade.ca

AI-powered financial analysis for Canadian investors. Deep research on 11,000+ stocks using a custom architecture that runs entirely on local hardware.

4× RTX 3090s, $0 cloud AI costs

Custom Ralph Loop architecture built because the cloud bill for equivalent AI processing would have been prohibitive.

◆ Lore & Knowledge

Lessons from 25 years of building real products — no email required, no pitch at the end.

Architecture

How I Run Multiple $10K MRR Companies on a $20/Month Stack

Single binaries, SQLite, and zero cloud vendor lock-in. The real cost of modern infrastructure isn't the bill — it's the complexity you take on.

Read on my blog →

AI Strategy

A Ralph Loop for Reading: Beating GPT with a 4K Context Window and 4 GPUs

Deep financial research on 11,000 stocks. Local GPUs instead of cloud APIs. The architecture that made it work when the obvious approach was too expensive.

Read on my blog →

Leadership

I Learned Mandarin. Here's What It Taught Me About B2C SaaS.

The learning curve for language and the learning curve for building products are the same shape. What the early wins teach you — and what they don't.

Read on my blog →

What Clients Say

"I needed someone with a sharp perspective on how to think about some of the architecture and critical components of my voice AI technology stack. Steve Hanov gives me that assurance. I'm lucky to have met him at Builders Club, now I have a heavy hitter in my corner when I need to make critical business decisions about complex engineering problems that are way out of my league."

— Ian Timotheos-Pilon, Voice AI Founder

◆ Player Profile

I've worked across the whole stack for 25 years — from radio protocols to SaaS products. I was an indie founder for 16 of them.

Steve Hanov

Senior Technical Consultant

The arc

I led EDGE/LTE firmware work at BlackBerry, built cellular protocol stacks, worked in the Office of the CTO, and filed five patents. Before that, I worked at Microsoft on IPv4/IPv6 networking. Then I spent 16 years as an indie founder, building and running profitable SaaS products without venture funding — WebSequenceDiagrams, Zwibbler, RhymeBrain, and eh-trade.ca.

Today I consult with companies and founders on AI strategy, embedded systems, software architecture, SaaS product strategy, technical due diligence, GTM, and rescuing over-complicated systems. I know what it feels like to be the person making the call — because I've been that person.

M.Math Computer Science, UW

Ex-BlackBerry (Office of the CTO)

Ex-Microsoft (IPv4/IPv6)

5 Patents (EDGE/LTE)

3D printing enthusiast

Mandarin speaker

200+ technical blog posts

11.5k Stack Overflow rep

Blog → | LinkedIn → | GitHub →

Let's Talk

If you need someone who can go from low-level technical details to product strategy in the same conversation — and tell you honestly what's what — send me a message.

What happens next

You send me a message. I reply within 24 hours — usually faster. We schedule a 30-minute call. No pitch, no pressure. You describe the problem. I'll tell you honestly whether I can help.

If it's not a fit, I'll point you to someone who might be. If it is, I'll outline a concrete next step — usually a lightweight diagnostic before any major commitment. You stay in control the whole way.

Email: steve.hanov@gmail.com

Location: Waterloo, Ontario, Canada

Remote: Yes — I work with teams worldwide

Engagements typically start at $15,000 USD.

Prefer to schedule directly?