ZAKHOMUTHContact

Writing

Substack

2026

2025

2024

2023

  • Age of Automation

    Businesses will hire accountable, configurable software services in place of employees and tools; toolmakers should sell the service, not the tool.

2018

  • And away we go… Merge!

    · Upverter blog

    Previews the Upverter and CircuitMaker merge under Altium: shared community, expanded parts library and desktop editing.

2017

  • Upverter Joins Altium

    · Upverter blog

    Announces Altium's acquisition of Upverter, the removal of paid tiers, and the plan to keep Upverter free.

  • Announcing EE Concierge

    · Upverter blog

    Launches EE Concierge, a paid service that makes verified parts for engineers who do not use Upverter, with stats and pricing.

2016

  • Whats Next at Upverter

    · Upverter blog

    Explains the quiet before Upverter version 4 and a shift toward plugins for users of desktop ECAD software.

2015

  • The Future of Circuit Design

    · Guest

    Why circuit design is moving from offline desktop tools to collaborative, cloud-based tools, with three common objections addressed.

  • Introducing Upverter 3.14

    · Upverter blog

    Announces Upverter 3.14: Parts Concierge, verified parts, constraints manager, system designer, on-premise enterprise and free manufacturing credit.

  • Disrupting Electronic Design Automation

    · Guest

    Four missing pieces that make EDA hard to disrupt: an open file format, a parts library, a manufacturing API, and an interchange language.

  • The Cost of Mistakes on Parts

    · Upverter blog

    Breaks down what a part-design error costs: re-spins, rework, debugging time, expedited boards and delayed time to market.

  • Why EDA Component Creation is Broken

    · Upverter blog

    Argues that hand-creating schematic symbols and footprints is tedious, error-prone and overdue for automation and shared data.

  • Creating the Perfect Part (4/4)

    · Upverter blog

    Final part of the series: footprint dimensions, pad sizing and solder-mask choices that keep a design manufacturable.

  • Creating the Perfect Part (3/4)

    · Upverter blog

    Part three of the series: footprint tips for reuse and readable silkscreen, with fabrication advice from Hooman Javdan.

  • Creating the Perfect Part (2/4)

    · Upverter blog

    Part two of the series: guidelines for readable schematic symbols, from splitting multi-unit parts to pin naming and placement.

  • Creating the Perfect Part (1/4)

    · Upverter blog

    Part one of a series on creating parts correctly, covering which attributes a new component library entry should include.

2014

  • Summer Features Blitz

    · Upverter blog

    Five new Upverter PCB editor features: layer translucency, selection filtering, guided routing, blind and buried vias, and keepouts.

  • Celebrating the Last 12 Months at Upverter

    · Upverter blog

    A look back at Upverter's fourth year, with an infographic of the past twelve months of features and figures.

  • On CircuitMaker and open-source EDA

    · Guest

    A Hacker News comment laying out Upverter's free, open-source, cloud-based position against Altium's CircuitMaker announcement.

  • Crowdrooster Interview

    · Upverter blog

    An interview with Crowdrooster, a social store for hardware after crowdfunding, covering its model, approval criteria and vision.

  • Picking the Right Bluetooth Module

    · Upverter blog

    A comparison table of popular Bluetooth modules, from price per unit to receiver sensitivity, for choosing one for a project.

2013

  • Disruption: Hacking Startup Travel

    Running a company from Toronto forces the CEO to travel. How to concentrate that disruption on one gritty person and keep the team productive.

2012

  • 25 Hardware Startups That Will Exist

    A 2012 list of twenty-five hardware problems startups were expected to solve, and a list of things that will look silly in hindsight.

  • Toronto Is Broken

    Toronto's startup ecosystem is young and broken: weak founder network, small ideas, no scale-experienced talent. Which parts to fix.

  • My Coach & My Scorecard

    An advisor's no-nonsense CEO scorecard: strong on vision, financing and recruiting, weak on marketing and evangelism.

  • Chateau Upverter

    Upverter's Toronto hacker house, the family-style culture behind it, and why early-stage founders should live and work together.

  • Build Your Own Hardware

    Every large software company builds hardware. The next giant in either category will have to own the whole user experience.

  • Sad, Tired, and Alone: My Ongoing Battle With Startup Depression

    A candid account of recurring founder depression, why it happens, and how to climb out.

  • Toronto: Dear Waterloo, It's Not A Zero Sum Game

    The Toronto/Waterloo startup rivalry is pointless. The future is equalization across cities, not a new single mecca.

  • Why The F@*K Didn't We Just Make An App?!?!

    Back-of-envelope app-economy math: software has become a commodity, and tooling is about to make hardware as easy to build as apps.

  • Toronto: How We Hire The Best.

    Upverter's recruiting playbook for Toronto: build a co-op pipeline early, earn referrals through brand and honesty, own the talent problem.

  • Toronto: Why Are We Here?

    Why Upverter stayed in Toronto after YC: the talent is there and wants to be there, even though it is a poor place to raise money.

2011

  • Why don't we have fully three dimensional integrated circuit layouts?

    · Quora

    A short Quora answer: ICs are designed as stacked 2D layers because that is how deposition and etching manufacture them.

  • Insight

    Two design goals for winning at community building: build something beautiful enough that people leave the tab open, and give them a reason to hit refresh.

  • Upverter in 130 Seconds

    The investor pitch: a $4.6B EDA market, incumbents that do not understand the web, an underserved hobbyist segment, GitHub and Salesforce as models.

  • Operational Transforms

    A short primer on operational transforms as the mechanism behind real-time collaborative editing, and why a schematic editor needed them.

  • My US Border Nightmare

    Detained and refused entry at the Lewiston bridge and then Pearson on the way to California, with a ranked list of the author's own mistakes.

2010

  • Parallax

    A visit to Parallax, makers of the BASIC Stamp, their decades of de facto open hardware, and the resolve to build tools for companies like them.

  • December?!?! Already?!?!

    Why a made-up mid-December demo deadline was set: it forces trimming to the critical problems and lines up with when fundraising must start.

  • Social Black Magic

    What the founders learned about company social presence: Twitter for reach, Facebook of questionable value, the blog as the only owned channel.

  • Reaching Around

    Cold-emailing prominent open-source-hardware people, and the takeaway that anyone notable should make their email and location easy to find.

  • In Browser, Pure Javascript, XMPP Client

    Why a browser chat client reusing Google and Facebook accounts needs its own XMPP server, a comparison of jabberd2, openfire, tigase and ejabberd, and a recommendation.

  • Interperting Ads-Words Results

    Google AdWords search volume as an order-of-magnitude market-size signal, in the Steve Blank customer-development sense.

  • Something Is Getting Harder

    A framework for early-stage security: address social threats socially, harden public interfaces to the 90% mark, audit on a sliding scale.

  • Backing Up Our Future

    Amazon S3 as a zero-maintenance backup target for the company's database and repos, with setup steps.

  • Split Personality

    Switching between business-development mode and hacking mode in one day wrecks both. Batch days by mode.

  • Zero (The Not-So Long Wait Is Over)

    Running out of personal money two months into full-time work, and reasoning about delaying fundraising until there is more to show.

  • User Experiance and Fist Over IP

    A UX crash course (Krug, Norman, Fowler) distilled: test on real people early, keep it self-evident, users scan, make the user feel smart.

  • Proof Of Concept

    Resisting the urge to build the whole tool, and instead proving the risky object-model hypotheses with a functional subset first.

  • Koolaide In The Ocean

    Blue Ocean Strategy, why investors respond to it, and the struggle of positioning a new market while still supplying conventional numbers.

  • Cost Effective Housing

    Line-by-line math on whether the founders should rent together in Toronto. A lease costs $20-56K of runway, so it waits until money is raised.

  • Coolest Font Ever

    How the name Upverter was found (an EE clock-multiplier term), why it worked, and the open-source Orbitron font in the first logo.

  • Whats In A Name?

    Ranked criteria for naming a web startup: spellable, .com available, verb-able, under eight characters.

  • Mantella VP

    A first VC networking event in Toronto: what worked and what to fix.

  • Glorious Metrics

    A homegrown Django metrics app tracking lines of code, check-ins, runway and update cadence, with the architecture behind it.

  • Django

    Evaluating Django against Rails, and noticing the community is large but light on professional programming experience.

  • Our Python Toolset

    The Python server-side toolchain: git, SCons, IPython, nose, coverage, PyLint and cyclomatic complexity scripts, with install and run commands.

  • Where is the Line?

    Architecture versus shipping: every design discussion must produce code within a few days.

  • Damn, This Boy Can Code

    The first 1,764 lines of Python, and why a Redis-style OBJ:ID:ATTRIBUTE key-value store beats MySQL for schemaless data.

  • Looking At The Competition

    Why OrCad, ViewDraw, KiCad, Eagle and gEDA all fall short: connections are hard to see, too many clicks, bad libraries, ugly, unhackable.

  • The Mystical Business Plan

    Skipping the formal business plan for an executive summary, and what is exciting and nerve-wracking about the market.

  • The End of an Era

    Why he left Sandvine: a company that aged into nine-to-five while the young unmarried engineers shipped every crisis.

  • Biz School

    A long brain-dump of everything a first-time founder does not know about VC terms, valuation, boards, hiring, PR, monetization and culture.

  • Setting Up Unit21.ca

    Every server-side tool chosen for the first company toolbox (web server, CMS, mail, LDAP, tracker, wiki, CRM, git, DMS, ERP) and why.

  • Resumes Suck

    Resumes strip out personality and group fit, the signals that matter most in hiring.

  • The Tribal Village

    Humans evolved for fifty-person villages. Which tribal role (leader, council, worker, misfit) do you naturally fill?

  • Cost of a Day in Life?

    Prospective budgets fail because they are contrived; retroactive, trend-based budgeting works, and a founder can live on about $10k a year.

  • Day One

    The decision to quit a job and start a company, told alongside the door-to-door computer business run at sixteen.