New startup ideas · AI and software · AI infra and compute

startup idea

Wattlot

Small businesses with spare power host inference pods; we hold the permits.

Wattlot is a marketplace where small industrial businesses with unused electrical service and yard space, machine shops, grain handlers, cold stores, rural ISPs, host containerized inference pods and get paid for delivered compute hours.

3/5

venture judge

9

similar startups, last 2 years (33 all-time)

99%

of 4 nearest real companies still alive

yes

8 matching federal grants and programs

Direction supported by government programs and grants

Test it before you build it

$1,800 · 5 weeks · 25 prospects

For under $2,000 and five weeks, prove that small industrial owners with spare 480V service will hand over utility bills and put down $500 refundable reservations to host a paying inference pod in their yard.

1Focus group: who and where

Owner or general manager of a machine shop, grain elevator, or cold storage business with 10 to 50 employees in rural electric cooperative territory in Iowa or Minnesota, with 480V three-phase service sized for peak loads they no longer run and a monthly demand charge they complain about

where to find 25 · LinkedIn Sales Navigator filtered to owners and GMs of machining, grain handling, and cold storage companies with 10-50 employees in Iowa and Minnesota; National Tooling and Machining Association local chapter meetings; GEAPS (Grain Elevator and Processing Society) member network; WISPA member list for rural ISPs; and the IMTS show floor in Chicago in September 2026 where thousands of shop owners walk the aisles

2Sell first, build later

A first-cohort hosting agreement sold before any pod is built: Wattlot files the interconnection pre-application and permits, commissions one containerized pod on the owner's slab within 120 days of site approval, and pays a guaranteed floor of $1,500 per month plus per-hour upside for 24 months

the ask · $500 refundable site reservation now; at commissioning, a $2,500 one-time permitting and commissioning fee netted from the first payouts, plus a 20% take rate on gross hosting revenue

a real yes · A real yes is a $500 deposit paid through the Stripe link and 12 months of utility bills in hand. Not a yes: 'sounds interesting, call me after harvest', a waitlist email with no bill, or a co-op that likes the idea but will not put an introduction in writing.

3Small experiments

The first one attacks the riskiest assumption; each ends with a number that says whether to run the next.

  1. 1. Twenty spare-capacity owner calls

    $200 · 14 days

    Call 20 owners with a one-line pitch: 'You are paying demand charges on service capacity you do not use; we place a containerized compute pod in your yard, handle every permit and utility filing, and pay you monthly per delivered compute hour.' Ask for 12 months of utility bills to size the opportunity; the bill share is the commitment signal, not the compliment.

    keep going if · 8 of 20 owners send utility bills within a week of the call

  2. 2. Payout calculator landing page

    $600 · 21 days

    Put up a one-page site with a hosting payout calculator (enter service size and average load, see estimated $1,500 to $4,000 per month per pod) and a waitlist form that requires a utility bill upload to join. Drive traffic with $400 of Google Ads on demand-charge and commercial-electric-bill search terms geo-fenced to Iowa and Minnesota, plus direct links sent in the call follow-ups.

    keep going if · 10 bill uploads from 40 form starts; a waitlist signup without a bill counts as nothing

  3. 3. Co-op channel pre-check

    $100 · 21 days

    Book calls with member services or key accounts managers at 5 Iowa or Minnesota electric distribution cooperatives (found via the NRECA co-op directory) and ask two questions: would aggregated behind-the-meter compute load on member sites need a new rate class or filing, and would the co-op introduce Wattlot to commercial members if it enrolls the pods in their load management program. This tests the stated regulatory risk and the channel in one pass.

    keep going if · 2 of 5 co-ops say existing commercial rate tariffs cover it and 1 agrees in writing to introduce members after a reference site exists

  4. 4. Reservation deposit close

    $100 · 14 days

    Offer every owner who shared bills a first-cohort reservation: $500 refundable deposit via a Stripe Payment Link that buys a site electrical survey, the interconnection pre-application filed with their co-op, and a locked payout rate for 24 months. Run the close on a 20-minute call with a two-page term sheet.

    keep going if · 5 of the bill-sharing owners pay the $500 deposit

4Collect a deposit up front

Tesla took $1,000 refundable reservations for the Model 3 and $100 for the Cybertruck before building either: the deposit is the measurement, not the revenue.

$500

per prospect, refundable

how · Refundable reservation via a Stripe Payment Link, paid by the business owner after signing a two-page reservation term sheet; the deposit converts to a credit against the $2,500 commissioning fee set up: Stripe Payment Links

what it reserves · One of the first ten pod slots, a completed site electrical survey, the interconnection pre-application filed with their co-op, and a payout rate locked for 24 months

refund · Refunded in full within 5 business days if the site fails the electrical survey, if the co-op filing is denied, or on request any time before pod delivery.

target · 5 paid $500 deposits from 20 owner conversations within 35 days

Go: build it if

5 deposits paid, 8+ bill shares, and at least 1 co-op confirming existing tariffs cover the load with a written willingness to introduce members; that proves supply will sign before hardware exists and the permitting wedge is real.

Kill: stop if

Fewer than 3 bill shares after 30 owner conversations, 0 deposits after 10 closes attempted, or 4 of 5 co-ops saying aggregated compute load needs a new rate case; any of those means the supply side or the regulatory wedge is not there.

Would you run this test?

One tap. The yes-share feeds the Demand pillar of this idea's score; nobody sees who answered.

Budgets are out-of-pocket estimates for a team of one to three, US market. Size the deposit to the deal, and check the terms before taking money in a regulated line.

Scorecard

One score that balances how trendy the idea is, the demand for it and its potential for 100x, with competition measured relative to every other idea in the catalog. Recent startup trends first, government priorities second.

71

Idea Score, 0-100 · raw 43.5 x 1.64

Active

competition: more crowded than 69% of ideas · headwind x0.65

+3.1

government priorities, secondary (22 matching grants)

Trend

72

Is the wave forming now? 2025-26 entrants vs 2023-24, rounds since 2025, the sector's live-batch direction, the 2026 trend analyst.

  • Entrants 2025-26 vs 2023-24 (similar companies)68
  • Rounds announced 2025+ in the sector68
  • Sector direction (live batch)100
  • 2026 trend analyst50

Demand

52

Does anyone want it? YC's current RFS, companies already paid for something similar, the operator judge, founders' yes-rate in decks, readers who would run the test.

  • YC asks for it (current RFS: idea / sector)30
  • Someone already pays (similar companies, recent / all-time)100
  • Operator judge: real pain25

100x potential

64

Can it return a fund? The venture judge (double weight), market-size and moat axes, neighbours still alive, the technologist judge.

  • Venture judge50
  • Market size axis100
  • Moat axis80
  • Neighbours still alive30
  • Technologist judge75

Score = 100 x cbrt(Trend x Demand x 100x) x (1 - 0.5 x crowding) + government bonus (max 5), calibrated so the 95th-percentile idea scores 90 (order never changes). A geometric mean: a weak pillar cannot be papered over. Percentiles are among the 272 ideas in the catalog; the terms matched were spare, power, host, inference, pods, hold, marketplace, industrial.

The idea in full

What
Wattlot is a marketplace where small industrial businesses with unused electrical service and yard space, machine shops, grain handlers, cold stores, rural ISPs, host containerized inference pods and get paid for delivered compute hours. Wattlot owns the hard part: the interconnection agreements, air and noise permits, utility demand-response enrollment and NERC-adjacent compliance that a single small business cannot obtain, plus the pod hardware and thermal design. Hosts are the paying side, giving up a take rate on payouts and a fee for the permitting and commissioning package.
Why now
Expanse (YC X26) is already monetizing wasted GPU capacity and Zibra Labs (YC X26) is building a distributed runtime for AI workloads at scale, so the software to use scattered supply exists; YC's Fall 2026 Compute at Sea request and Atomarine's floating nuclear data centers show how far builders will go for interconnected power, while the cheapest unused megawatts sit behind small business meters on land.
Wedge: first customer and entry point
Rural electric cooperative territory in one state: sign twenty existing commercial members with 480V service and idle load, using the co-op as the channel and the single permitting jurisdiction as the template.
Path to 100x
Data center capacity is a $100B+ annual build, and every gigawatt Wattlot assembles from small business sites is capacity that never enters a five-year interconnection queue. The regulatory moat compounds: permit templates, co-op agreements and demand-response enrollments per jurisdiction are slow to copy, and once a state's small business power is contracted, the second entrant finds no supply left.
Ceiling
Pod economics stay hostage to hardware depreciation and site-by-site permitting cost, so the business scales linearly with field crews instead of with the network.
Closest real companies, as the generator saw them
Expanse unlocks wasted GPU capacity that already exists; Wattlot creates supply by placing hardware behind small business meters, and its defensibility is the permit and interconnection stack rather than the scheduler. Zibra Labs and OpenRelay supply the hardware-agnostic runtime that makes such fragmented capacity sellable.
Main risk
Utilities and state regulators classify aggregated behind-the-meter compute as a new load class and stall approvals for years.

Five judges

Each judge scores every idea in the catalog with a named rubric; the venture judge decides whether a card is shown at all (4-5 is venture-grade).

  • Venture investor

    3/5

    Permit stack per jurisdiction is a real moat, but the card admits pod economics may scale linearly with field crews and depreciation.

  • Bootstrapper

    1/5

    Pod hardware, interconnection agreements and per-jurisdiction air and noise permits burn far past $500k before one compute hour is billed.

  • Operator

    2/5

    The grain handler with idle 480V service pays a commissioning fee for a benefit that arrives only after permits clear, which is the wrong side paying.

  • Technologist

    4/5

    Pod thermal design plus per-jurisdiction interconnection and NERC-adjacent permit stacks is slow, physical expertise that a scheduler company cannot copy.

  • Risk

    3/5

    Permits, interconnection agreements and co-op contracts are genuine licensed footing, but one state regulator creating a new load class stalls approvals for years.

  • trends

    3/5

    Same 2025-2026 power scarcity as Swingbus, but multi-year site-by-site permitting means the window likely closes before behind-the-meter pods reach scale.

Similar startups in the directory

Companies whose pitch matches most of the idea's terms (spare, power, host, inference, pods, hold, marketplace, industrial): 33 all-time, 9 from the last two years. Same matching as Idea Check.

  • Cortex AIyc F25 · 2025 · Data for AIalive

    Large-scale real-world robot & human data for embodied AI

  • Terra Bioforgeplugandplay PnP 2026 · 2026alive

    Terra Bioforge engineers overproduction of microbial-based natural product compounds with billion-dollar end markets.

  • Riftenyc S26 · 2026 · AI infra and computealive

    Earned intelligence for every company

  • Gazzeboplugandplay PnP 2026 · 2026 · Commerce and marketplacesalive

    Gazzebo is an AI-powered procurement marketplace modernizing how corporate travel buyers discover, evaluate, and engage managed travel suppliers.

  • Tinfoilyc X25 · 2025 · AI infra and computealive

    Encrypted AI with verifiable privacy

  • atronous.aiplugandplay PnP 2025 · 2025 · Commerce and marketplacesalive

    Automating product digitization and marketplace submissions to drive efficiency and higher conversion rates for retailers and suppliers.

  • Axiom Therapeuticsplugandplay PnP 2025 · 2025 · Healthcare and bioalive

    AI-powered covalent drug discovery for untreatable cancers

  • Laibltechstars TS 2025 · 2025 · Robotics and physical worldalive

    Laibl — White-Label Marketplace Technology

  • Malty Snacksplugandplay PnP 2024 · 2024 · Consumeralive

    We are a food tech company with a mission to transform the food industry through our upcycled ingredients and innovative snacks.

  • Ubicloudyc W24 · 2024 · AI infra and computealive

    Open source alternative to AWS

  • Garageyc W24 · 2024 · Commerce and marketplacesalive

    Marketplace for America's most essential assets

  • BeCausealchemist Alchemist Class 33 · 2023 · Climate and energyunchecked

    AI-Powered Sustainability Management Hub for the Tourism Ecosystem

Run this as an Idea Check →

The generator's reference companies

Real companies the model named as closest when it wrote the card, with their fate. A check mark is a company the radar could verify in its directory.

Public money in this direction

US federal grants, SBIR/STTR awards and open opportunities from the radar's public-money feed, matched to the idea's terms; the sector totals give the context.

8

grants and programs matching the idea

225

startup-relevant grants in AI infra and compute

$271M

awarded in the sector, tracked

8

opportunities open now in the sector

All public money by sector →

Market signal

What the radar sees in AI infra and compute: new companies by cohort year, the forming YC batch, and outcomes since the February snapshot.

AI infra and compute · 10 → 33 → 57 → 61 → 52 new companies 2022 → 2026 · 98% aliveYC S26: 18 in this cluster, 8% of the batch (was 4% in X26) (F26 is still forming: 21 listed)Since February, of 117 YC companies here: 3 acquired, 2 shut down, 25 rewrote their pitch

AI infra and compute: companies, trend and grants →

Design attributes

The card is one cell of a designed set: every axis below was chosen before the text was written, and the text had to realize it.

Buyer
Small business
Business model
Marketplace
Path to 100x
Creates a new category
Market size
$100B+ market
Capital intensity
Capital-heavy (hardware, bio, infra)
Speed to revenue
Revenue in 1-3 years
Technical depth
Deep tech: ML, hardware, bio
Go-to-market
Partners and channels
Moat
License or regulatory moat
Geography
US first
Regulation
Heavily regulated
Vibe
Boring business

Listed under

An idea sits in its own sector and in any sector its text clearly touches.

More ideas like this

AI and software · AI infra and compute

Residua

Secondary market where enterprises resell unused committed cloud and GPU spend.

Residua is a self-serve exchange where enterprises list the unused portion of multi-year committed spend contracts with clouds and GPU neoclouds, and other enterprises buy that entitlement at a discount.

Score 98Warm competitionVC 4/5MarketplaceEnterprise99% of 4 neighbours alive

AI and software · AI infra and compute

Auralith

Over-the-counter hearing wearable with a speech model burned into custom silicon.

Auralith sells a consumer ear-worn device whose noise separation and speech reconstruction model is hardcoded into a low-power ASIC, so a full conversational model runs for a day on a coin-cell class battery with no phone and no cloud.

Score 87Active competitionVC 4/5HardwareConsumertest: $1.5k · 5w99% of 4 neighbours alive

AI and software · AI infra and compute

Gridshift

The marketplace where AI data centers sell their flexibility to the power grid.

Gridshift turns interruptible AI training load into a tradable grid product: data centers enroll clusters, Gridshift's ML forecasts and checkpoint-aware orchestration shift or shed training jobs on minutes of notice, and the flexibility is sold into wholesale electricity markets as demand response.

Score 74Crowded competitionVC 5/5MarketplaceEnterprise99% of 4 neighbours alive

AI and software · AI infra and compute

Backline

Voice and back-office agents for small businesses, running on our own inference fleet.

Backline answers the phone, books jobs, chases invoices and updates the booking system for plumbers, clinics, garages and salons, billed per handled call with no software for the owner to configure.

Score 72Crowded competitionVC 4/5AI agent as a serviceSmall businesstest: $1.5k · 5w99% of 4 neighbours alive

AI and software · AI infra and compute

Ledgerwatt

Metering and settlement API that turns AI spend into an auditable clearing layer.

Ledgerwatt sits between enterprises and every model provider, neocloud and GPU broker they use, meters each call at the token and job level, prices it against the contracted rate card, and settles a single invoice with per-team, per-product attribution that finance can audit.

Score 71Crowded competitionVC 5/5Infrastructure and APIsEnterprisetest: $1.2k · 6w99% of 4 neighbours alive

AI and software · AI infra and compute

Innervox

A neural wristband that turns silent speech into text for AI assistants.

Innervox builds a consumer wearable that reads surface EMG signals from subvocalized speech, letting people talk to AI assistants silently at conversational speed - in meetings, open offices, or public spaces.

Score 71Active competitionVC 4/5HardwareConsumer99% of 4 neighbours alive
Swipe ideas like this in the deckTalk to the radar about it

Fictional company written 2026-08-22 from MarkosWeb data; the companies, grants and numbers around it are real and tracked. Treat the idea as a research prompt, not a plan.