Frontier Report 001 · Version 1.0

The first 70 days

Software became abundant faster than evidence.

What the first Parallel Human portfolio supports, what it does not, and what must be measured before PH can make claims about company formation, founder capacity, or capital.

Executive finding

The production layer changed. Company formation remains unproven.

During the opening Founder 0 case, one operator without a prior software-development career directed a broad set of venture builds using machine assistance. By July 15, the canonical PH portfolio contained eleven ventures across five evidence stages. Some were live, some had design partners, one services operation was paid, and others remained prototypes or MVPs.

That is meaningful evidence of expanded software-production capacity. It is not evidence that eleven companies were created in seventy days. At least three sampled portfolio code histories predate the May 5 observation boundary. The record also lacks independently verified labour, full economic cost, a conventional comparison, governed prospective founder-week data, and repeatability across founders.

Report 001 supports the question. It does not support the mythology.
ObservedBroad product output

Eleven ventures appear in the canonical public snapshot at different stages.

MixedProduction compression

The output is real, but equivalent labour, quality, and cost comparisons are missing.

Not testableCompany formation

Durable demand, economics, distribution, and operating continuity are not established across the portfolio.

Not testableCapital advantage

Receipt reconstruction and a valid conventional comparison do not yet exist.

The dataset

Eleven ventures, five stages, and uneven proof.

The report uses the public portfolio snapshot dated July 15, 2026. Stage describes current evidence maturity. It does not describe valuation, investment readiness, portfolio allocation, or future potential.

11ventures in the canonical snapshotObserved public record
5evidence stages representedPrototype through Paid
2public product URLs in the snapshotPublicly inspectable links
1Founder 0 caseNo external comparison group
StageCountWhat the stage establishesWhat it does not establish
Paid1A paid services relationship is recorded.Repeatable revenue or software economics.
Production3The product is used or publicly live.Product-market fit or durable distribution.
Design Partner2A named relationship and usable product are recorded.Paid adoption or scalable demand.
MVP3A bounded product exists.External use or commercial evidence.
Prototype2The idea has working form.Customer evidence or venture viability.

Important correction: the portfolio count is a current snapshot, not a creation count for the observation window. Sampled path-specific repository history shows at least three portfolio code histories began before May 5. Eight sampled paths show first tracked activity inside the window. A first commit is an activity marker, not a company-formation date.

Method

A bounded reconstruction, not a controlled comparison.

Report 001 combines four evidence sources:

  1. Canonical portfolio snapshot. The July 15 venture record supplies names, stages, public-safe proof statements, and public links.
  2. Sampled repository history. Path-specific Git histories across the eleven mapped venture codebases or directories supply first tracked activity inside or outside the observation window. Commit volume is excluded because repository structures differ and commits are not equal units of work.
  3. Public artifacts. The report treats only URLs present in the canonical snapshot as publicly inspectable product links.
  4. Operator estimates. The sixty-hour week and CAD $1,500 direct-spend figures are retained as estimates reported on July 13. They are not treated as verified cost findings.
Included

Venture stage distribution, public snapshot claims, sampled tracked-activity dates, protocol status, and explicit missing evidence.

Excluded

Private customer details, unverified revenue amounts, health data, session transcripts, model-generated time estimates, and invented conventional build costs.

The report is retrospective. PH's protected prospective Founder 0 observation store has not been approved. No post-July 15 founder-week or venture-week observations are used.

Findings

Five things the record can say now.

01 · Mixed evidence

One non-developer can direct much more software production than before.

The portfolio contains functioning products, live public surfaces, design-partner work, and a paid services operation. The code histories show sustained technical output across several categories. This is stronger than a demo claim.

The missing comparison matters. PH has not reconstructed equivalent specialist labour, review time, rework, shared infrastructure, or quality criteria. Report 001 therefore records directional production compression, not a measured multiplier.

02 · Observed boundary

External evidence stayed scarce after software became easier to produce.

Only two public product URLs appear in the canonical snapshot. Other proof depends on operating or relationship records: daily use, design partners, a paying client, testing status, or no external evidence yet. That does not make the claims false. It makes their verification different and less public.

The portfolio therefore grew faster at the production layer than at the public evidence layer.

03 · Counterexample to the headline

The eleven-venture snapshot is not an eleven-venture creation event.

At least three sampled portfolio histories began before the observation window. Several ventures also emerged from prior operating relationships, existing user knowledge, or work already underway. The correct claim is that the portfolio was advanced and made legible during the window, not that every venture began from zero after May 5.

04 · Working interpretation

The bottleneck moved into human judgment and external consequence.

When software output expands, problem choice, trusted access, customer truth, product taste, sales, distribution, compliance, and attention become more visible constraints. The current record supports this as an operating interpretation, not as a causal finding.

05 · Research finding

Portfolio count is a poor measure of founder capacity.

A paid service, a regulated design-partner product, a public-good platform, and an early consumer prototype create different human loads. Counting all four as one venture each hides the actual work. PH must measure founder-weeks, venture-weeks, obligations, review, correction, external meetings, and dependency class before making a capacity claim.

Limits and counterevidence

The strongest criticism belongs inside the result.

Pre-window work

At least three sampled venture histories began before May 5. The window does not represent a clean start.

One selected operator

Founder 0 brought sales, strategy, operating experience, product judgment, and existing relationships.

Output is not traction

Eleven venture records do not mean eleven proven companies, durable markets, or investable outcomes.

Hidden labour

Direction, review, correction, research, sales, and context switching are not fully reconstructed.

Unverified spend

CAD $1,500 is an operator estimate. Invoices, shared tools, domains, model usage, and opportunity cost are incomplete.

No comparison group

No equivalent conventional team, matched project, or external founder dataset exists.

Evidence access

Several relationship and use claims depend on internal records that a public reader cannot independently inspect.

Retrospective bias

The account was reconstructed after much of the work occurred. Prospective measurement remains blocked.

PH-SP-001 status

Report 001 does not close a registered hypothesis.

HypothesisStateReason
H1 Production compressionMixedBroad production is observed. Equivalent labour, cash, and quality comparison is missing.
H2 Portfolio capacityNot testableNo governed prospective founder-week and venture-week series exists.
H3 Evidence-led concentrationNot testableThe registered 180-day decision record and allocation history do not exist.
H4 Later or smaller capitalNot testableFull cost and conventional capital comparisons are incomplete.
H5 Founder replicationNot testableNo external founder has entered the governed study.
H6 Venture transferNot testableNo eligible governed transfer set exists.

Implications

Useful conclusions without pretending the model is proven.

For founders

Use machine intelligence to lower the cost of learning, not to maximize the number of products you can claim. Track external evidence, obligations, review, rework, and attention. Concentrate when the evidence changes the decision.

Start a founder conversation

For capital partners

Do not infer that early companies need less capital from this case. Ask what capital buys after product formation: founder runway, trusted access, distribution, compliance, data, hardware, customer success, or scale. Demand full economic cost before accepting a capital-efficiency claim.

Open the capital-partner path

For Parallel Human

The next asset is not another venture count. It is a better evidence system: prospective records, cost reconstruction, governed decisions, public corrections, and a founder pilot designed around value and rights.

Inspect the protocol

What must happen next

Turn the opening case into a testable program.

  1. 01
    Approve a protected prospective store.

    Collect founder-week and venture-week observations without putting sensitive records in Git, chat, or public systems.

  2. 02
    Reconstruct full economic cost.

    Verify direct spend, shared infrastructure, human direction, review, correction, and opportunity cost.

  3. 03
    Publish evidence and decision lineage.

    Separate maturity from portfolio allocation and record why ventures focus, reserve, transfer, or close.

  4. 04
    Design the smallest useful founder pilot.

    Open only after the offer, support, participant rights, consent, privacy, and stop rules are clear.

  5. 05
    Publish Report 002.

    Use prospective evidence, a clearer comparison, documented counterevidence, and external challenge.

Public claim register

Every material statement has a state.

IDClaimStateEvidence
FR001-C01The July 15 public portfolio snapshot contains eleven ventures.ObservedCanonical venture JSON.
FR001-C02The snapshot contains one Paid, three Production, two Design Partner, three MVP, and two Prototype ventures.ObservedDerived from canonical stages.
FR001-C03At least three sampled portfolio code histories predate May 5.Observed, boundedPath-specific Git history inspected July 24, 2026.
FR001-C04Eight sampled portfolio paths show first tracked activity inside the observation window.Observed, boundedPath-specific Git history. First activity is not company formation.
FR001-C05Founder 0 worked about sixty hours per week.Operator estimateNo independent time reconstruction.
FR001-C06Direct spend was about CAD $1,500.Operator estimateInvoice reconstruction incomplete.
FR001-C07Machine assistance materially compressed production.MixedOutput is observed; equivalent cost, labour, and quality comparison is missing.
FR001-C08Parallel Human compressed company formation.Not testableDurable demand, economics, distribution, and comparison are incomplete.
FR001-C09Parallel Human ventures require less early capital.Not testableFull economic cost and conventional scenarios are absent.
FR001-C10The operating model can replicate across founders.Not testableNo external founder study exists.

Download the machine-readable report register

Sources, conflicts, and corrections

How to challenge this report.

Source register

  1. Parallel Human canonical public venture snapshot, updated July 15, 2026.
  2. PH-SP-001 public protocol summary, version 0.1.
  3. Path-specific local repository histories for the eleven mapped portfolio ventures, inspected July 24, 2026. Only aggregate activity-boundary findings are published here.
  4. F1 Fantasy Strategy Engine and Ramara Hub, the two public product URLs in the canonical snapshot at the evidence cutoff.
  5. Founder 0 baseline estimates recorded July 13, 2026.

Conflict statement

Russell Cole founded and operates Parallel Human, is Founder 0, directs the ventures, controls the source systems, and may benefit from a favourable public story. Version 1.0 received internal editorial and automated claim checks. It has not received independent external subject-matter review. Readers should treat that as a material limitation.

Data availability

The canonical public venture snapshot and machine-readable claim register are linked above. Private customer, commercial, repository, and relationship records are not published. Claims that rely on those records are labelled accordingly.

Corrections

Send a specific claim ID, evidence, and proposed correction to hello@parallelhuman.ai. Material corrections will be dated in this register and propagated to the machine-readable version.

July 24, 2026
Version 1.0 published. The earlier dossier was closed. “Eleven ventures” was explicitly bounded as a current snapshot, not an eleven-venture creation claim.
Current status
No published corrections.

Follow the next test

Get Report 002 and material corrections.

The evidence stays public. Email is for delivery, corrections, and the next report.