The Parallel Quarterly — Issue No. 3 Fund III · First close Filed from Pittsburgh & Rotterdam

The future is
mostly plumbing.

Parallel is a $120M pre-seed and seed fund for the companies that keep factories, grids, warehouses and labs running. We write first checks of $500k–$3M, decide in ten days, and answer our own email.

Send us the deck Read the thesis
Fund III, at a glance
Fund size
$120,000,000
First check
$500k–$3M
Stage
Pre-seed & seed
Decision
Ten days, in writing
Partners
Two (see Fig. 1)
Reserves
1 : 1
§ 01 — The thesisParallel Quarterly · p. 2

In praise of load-bearing companies.

By Hana Voss & Teo OkaforFiled September 2026

Every decade, venture capital discovers a new way to arrive early at the same party. Social. Mobile. Crypto. Whatever this year is called. The money moves in a flock, and the flock is loud.

We are not in the flock. We are in the building next door, where someone is trying to get a 1987 controller to talk to a cloud dashboard without setting anything on fire.1The controller is beige, runs ladder logic, and has outlived two owners and a roof. It is, we’d argue, the most important computer in the building.

Here is the thesis, plainly. The next thirty years will be spent rebuilding the physical economy: the grid, the factory, the warehouse, the lab. That work needs software written by people who have stood on a shop floor, hardware designed by people who have read an interconnection queue,2About two terawatts of proposed generation and storage are waiting for permission to plug into the U.S. grid. The median wait is roughly five years. This is a software problem wearing a hard hat. and companies patient enough to sell to buyers who take nine months to say yes.

These companies are hard to fund at the start. They don’t demo well. Their markets look small on a slide and enormous in a spreadsheet. Their founders tend to be forty-one, not twenty-three,3Median founder age at our first check, Funds I and II: 41. Median years in the industry before founding: 14. Median hoodies: not disclosed. and would rather talk about torque than TAM.

That is the opportunity. Not because unfashionable is a virtue — it isn’t — but because fashion sets prices. When everyone wants the same thing, it costs too much. When nobody wants a thing, you can buy it at a sensible price and wait for the world to notice it was load-bearing.

“Let me take it to the Monday meeting” is the most expensive sentence in venture. We don’t say it. There are two of us, we write the first check, and we decide in ten days — yes or no, in writing, with reasons.4Our no’s come with a paragraph. Founders tell us it’s the most useful thing an investor has ever sent them. One had it framed. We are choosing to take that as a compliment. Nobody gets handed to an associate. We are the Monday meeting.

We are patient, but not in the soft sense. We are patient the way a bridge is patient. We expect to hold for ten years, and we intend to be useful for all of them.

If that sounds slow, good. The grid was built slowly, too. It still works.

— H.V. & T.O., Pittsburgh, September 2026

§ 02 — What we backParallel Quarterly · p. 3

Four places the real economy is stuck.

We don’t say “deep tech” or “hard tech.” We back companies that sell to people in steel-toe boots. Four theses, rewritten every January and argued about every Tuesday.

I.Manufacturing software

The factory floor is the last great offline network.

Fig. 2Ladder logic, c. 1987. Two rails, still parallel, still running the plant.

There are roughly a quarter of a million manufacturing sites in the United States, and most of them run on software older than their newest hire. The machines are connected to each other the way a family is connected: closely, stubbornly, and without documentation.

We back tools built by people who have held a torque wrench — scheduling, quality, maintenance, and the unglamorous glue between a thirty-year-old controller and anything with an API. We do not back dashboards that need a new factory to work.

What we look for A founder who has worked a shift. A pilot in a plant, not a pilot program.

See the four companies
II.Grid & energy hardware

The grid is a century-old machine with a new job.

Fig. 3A substation, single-line. The red box is where our founders tend to show up.

Electricity demand is rising for the first time in a generation, and the equipment that moves it was designed for a quieter century. We back sensors, controls and power electronics that a utility can buy, install and — crucially — insure.

What we look for UL listing on the roadmap. A lineworker on the advisory board.

See the five companies
III.Logistics

Freight still moves on phone calls and PDFs.

Fig. 4Rotterdam to Pittsburgh. The expensive part is the part where nothing moves.

A container crosses the Atlantic in eleven days, then waits four at the port for a chassis. We back software that removes waiting: yards, docks, drayage, returns. Nobody posts about it. Everybody pays for it.

What we look for Revenue per load, not per seat. A dispatcher who swears by it — or at least stops swearing.

See the four companies
IV.Lab tooling

Science is bottlenecked by pipettes, not ideas.

Fig. 5Serial dilution, by hand, in 96 wells. Our founders would like a word.

A good biology lab can dream up a thousand experiments and run about forty. The gap is instruments, inventory and the clipboard taped to the freezer. We back tools that turn benches into systems without asking scientists to become software engineers.

What we look for A lab manager as the champion. Consumables in the business model. Nobody saying “platform” before revenue.

See the four companies
§ 03 — Portfolio indexParallel Quarterly · p. 4

The index.

Every company we’ve backed that is allowed to be named. Gaps in the numbering are companies in stealth — which, in our sectors, mostly means busy.

Showing 17 of 17

Parallel portfolio companies, Funds I–III
No.CompanyWhat they doSectorFirst checkYearStatus
01QuenchHeat-treat process control for tier-two auto suppliers.ManufacturingSeed2018Acquired ’24
02PalletierDock scheduling for warehouses that still run on whiteboards.LogisticsSeed2019Acquired ’23
04Sieve InstrumentsCalibrated bench instruments with an honest API.LabSeed2019Acquired ’25
05OhmlineSubstation controllers that utilities can actually update.Grid & EnergySeed2020Active
07EmptylegBooks the return trip that regional trucks usually drive empty.LogisticsSeed2020Acquired ’25
09TorquewellHears a CNC spindle failing three weeks before it does.ManufacturingPre-seed2021Active
10DraylinePort drayage without the forty phone calls.LogisticsSeed2021Active
12MeniscusLiquid-handling robots that fit on an ordinary bench.LabPre-seed2022Active
14KilnworksElectric industrial heat for glass and ceramics.Grid & EnergySeed2022Active
15HoistlineCrane telemetry for steel service centers.ManufacturingSeed2022Active
17AgarworkCounts cell colonies faster than a graduate student, and without complaint.LabPre-seed2023Active
20ClampwiseTransformer health sensors that install in eleven minutes.Grid & EnergyPre-seed2024Active
22TareweightWeigh-in-motion scales for yards and depots.LogisticsPre-seed2024Active
25RoutecardJob travelers for high-mix machine shops, finally off paper.ManufacturingPre-seed2025Active
27PhaseyardInterconnection studies in weeks instead of years.Grid & EnergySeed2025Active
29TitrateReagent inventory, tracked by the milliliter.LabPre-seed2025Active
31Redacted — company in stealth[redacted] software for [redacted] that we can’t discuss until October.Grid & EnergyPre-seed2026Stealth

Fourteen more companies are in stealth. They will tell you themselves, eventually, at great length.

Funds I–III · updated 24 Sep 2026
§ 04 — By the numbersParallel Quarterly · p. 5
8days

Median, first call to written decision · Fund II · n = 212The promise is ten. We would rather be early than eloquent.

Figures we’re willing to print.

  • 92%

    of founders we’d back again. The other 8% we’d back again with a CFO.

  • 41

    median founder age at our first check. Wisdom teeth: removed.

  • $1.4M

    average first check in Fund II, with the same again held in reserve.

  • 0

    associates, principals or “platform” hires standing between you and a partner.

  • 31

    companies across three funds. Four acquired, none by a SPAC.

§ 05 — Founders, on the recordParallel Quarterly · p. 6

Founders, on the record.

We asked three founders what working with us is actually like. We did not edit the answers, which our lawyers found brave.

“They said no in nine days, with two pages on why. I rebuilt the plan around those two pages. Eleven months later they said yes — in six.”

Priya RamanCo-founder & CEO, Phaseyard

“Hana has been on our shop floor more often than our last three investors combined. She knows where the good coffee is. It’s in quality control.”

Marco DeLucaFounder, Torquewell

“Every investor asked about our TAM. Teo asked about our scrap rate. That was most of the diligence, and it was the right question.”

Ruth AdeyemiFounder, Quench — acquired 2024
§ 06 — How we workParallel Quarterly · p. 7

Ten days, then ten years.

  1. Day0

    First call

    Forty-five minutes with one partner. Bring the plan, not the deck. Before we hang up, we’ll tell you what worries us.

  2. Day3

    Partner meeting

    Both of us, usually at your place, because we’d like to see the machine. One customer reference. No data room.

  3. Day10

    Decision, in writing

    A yes is a one-page term sheet on standard terms. A no is a paragraph you can actually use.

  4. Day100

    The first hundred days

    Three introductions to buyers we’ve sold to, one senior hire we help close, and a board seat only if you ask for one.

  5. Day3,650

    Still here

    Reserves at one-to-one, follow-ons without theatre, and the same two phone numbers.

§ 07 — LettersParallel Quarterly · p. 8

Letters from the shop floor.

Quarterly, long, footnoted. Written by the partners, edited by nobody, and read by an alarming number of utility executives.

  1. Every few years someone announces the end of the programmable logic controller. The controllers, being controllers, do not respond. A field note on building software that respects the beige box instead of trying to replace it.

    — Teo Okafor, from a machine shop in Erie

  2. Slow buyers are sticky buyers. The procurement office that takes three quarters to sign will take a decade to leave. Why we underwrite patience as a moat, and how founders survive the wait without selling a kidney.

    — Hana Voss

  3. The best company we ever backed showed us a spreadsheet and a burn scar. On why the infrastructure that matters most is usually the kind that looks worst on a projector.

    — Hana Voss & Teo Okafor

  4. The scarcest resource in American energy is not lithium or land. It is a place in line. A long letter about queues, studies, and the founders turning a five-year wait into a spreadsheet problem.

    — Teo Okafor

  5. Who signs, who blocks, who actually uses the thing, and why the answer to all three is sometimes the same man named Dale. Includes a glossary and an apology to Dale.

    — Hana Voss

  6. We sent 212 rejections last year, each with reasons. We asked the founders which reasons were wrong. Here are the eleven times we would take it back, and the one time we still wouldn’t.

    — Hana Voss & Teo Okafor

§ 08 — Correspondence · p. 9

Bring us the one they called too heavy.

Hardware in the bill of materials. Nine-month sales cycles. Customers who still fax. If three investors passed because it was “too industrial,” you are in the right inbox. Both partners read every submission.

The pitch form needs JavaScript. Email your deck, or a paragraph, to decks@parallel.fund. Both partners read it.

No NDA, no warm intro, no deck-format police. We reply within ten days either way — and if it’s a no, you’ll get the paragraph.

Received. Both of us will read it before Friday.

Your reference is PAR-3-0418. If it’s a yes, expect a one-page term sheet. If it’s a no, expect a paragraph worth keeping. Either way, you’ll hear from a partner, not a form letter.