outricher/Visa ICP playbook/Profiles/ICP 01
ICP 01 · The status clock, engineering edition · ordered 10×
Foreign degree, US job, seven years or less
profiles delivered
ordered, March to August
email coverage on shipped cuts
The anchor of the whole account — and the client named it, not us. A person who earned at least one degree outside the United States and now holds a US machine-learning or data role is, structurally, someone who moved mid-career. The cap on US tenure keeps out everyone who settled here a decade ago. This is the cohort where an O-1A or an NIW argument lands most easily, and the practice’s economics turn on it: the hands-on practitioner self-petitions and pays, where an academic waits for an employer to file.
Hypothesis
Foreign degree + US job + short US tenure = on temporary status, and aware of it.
Signal
In the US now ∧ AI/ML or data title ∧ 6–15 years of career, ≤7 of them in the US ∧ at least one degree from a non-US institution.
Proof
Every foreign-degree claim checked against a register of 7,650 US institution name variants; 30/30 matcher tests green; 100% geography and title compliance on every shipped file.
Five rows from a delivered file
Who is actually in it
Anonymized to role, employer size band, region and signal — no candidate, client or employer is ever named.
| Who | Where they work | Location | The signal | Reach |
|---|---|---|---|---|
| Senior ML engineer | ~110-person US tech company | United States | Indian IT-institute degree · 9 yrs total, 5 in the US | personal email ✓ |
| Senior ML engineer, computer vision | ~950-person film company | United States | Iranian technical-university degree · 6 yrs total | personal email ✓ |
| Lead GenAI engineer | ~40-person real-estate tech company | United States | Indian research-institute degree · 13 yrs total, 3 in the US | personal email ✓ |
| Senior data scientist | management consultancy, ~3,600 staff | San Francisco | Institute-of-technology degree · 10 yrs total, 6 in the US | personal email ✓ |
| Machine learning engineer | ~370-person employer | United States | Canadian degree · 0 years in the US — a brand-new arrival | personal email ✓ |
The volume
What was delivered, and from what
11,500 profiles across ten batches, March to August — email coverage 98.9–100% on every shipped cut. He calls it his working configuration and asks for it by name.
The June experiment described below, 7,188 rows: [CLARIFY — counted inside the 11,500 or not].
What this means for your practice
Open on the clock someone else controls: a recent arrival in a technical role, on status tied to an employer. The O-1A and NIW arguments both rest on work the person has already done, so the first call is about evidence they already hold — not about whether they qualify to think about it.
Where we got it wrong, or nearly did
The batch that failed
In June we tried to make the list more impressive — narrower experience window, a name-brand employer required in every history, a size gate on the current employer. 7,188 rows, fully audited — and his verdict, recorded when he placed the next order, was that the cohort barely connected and barely replied. Both gates were deleted and the previous configuration restored wholesale. Prestige survives only as a ranking nudge inside the score, never as a filter. He did not walk. He asked for the old setup back, and ordered four more times.
What Outricher does
Custom research on a career graph, sold per record.
Outricher builds custom research on a career graph — dated positions, education records, awards, patents, publications and a contact layer — and ships it as ranked, evidence-tagged cohorts. The visa practice described here is one client; the method is the same for any ICP you can put into words.
for a first sample of your own ICP — [SAMPLE SIZE, e.g. 25 rows] ranked rows within [SAMPLE TURNAROUND, e.g. 48 hours]
per contact, flat — never a share of your engagement fee. Pricing
profiles built for one visa practice. See them all
Want this cohort for your own practice?
Describe the version of it you want — origin, field, seniority, years in the US — and we run it against the graph. Back comes a free sample of real ranked candidates with the evidence attached and contactability resolved.
The other five profiles