06-reference/research

anthropic tiger team partner model

2026-06-24·research-brief·source: deep-research
anthropicpartnershipsphdatagtmpositioning

"Tiger Team" Is phData's Word, Not Anthropic's: What the Anthropic Partner Model Actually Requires of SIs

The question

What does Anthropic's Tiger Team partner model require of designated SIs — certifications, headcount commitments, case study types, revenue targets — and which data-consulting firms have publicly landed Tiger-status engagements in 2025-2026? Context: the phData career fork doc (2026-06-18) shows the Anthropic "Tiger Team" route still live with a ~July 2026 window for the founder to lock a first-pull commitment with Andrew.

What we already know (from the vault)

What the web says

Convergences and contradictions

Synthesis for RDCO

The most useful thing this brief does is dissolve a false premise before it drives a decision. There is no Anthropic "Tiger Team" partner designation, no published list of its requirements, and therefore no roster of "data-consulting firms that landed Tiger status." The founder's option #4 is a phData org-chart team Andrew wants to stand up, named internally after the SWAT/special-forces idiom. Whatever requirements that team operates under are phData's to define (the $10M-from-$0 revenue goal is phData's number, not Anthropic's), layered on top of phData's firm-level standing in Anthropic's Services Track. So the actionable question is not "what does Anthropic require of Tiger partners" but two separable ones: (a) what does Anthropic's Services Track require of phData as a firm — answered and stable: certs + production customers + public stories, no revenue or headcount gates; and (b) what does Andrew require of a first-pull hire onto his team — answerable only by Andrew, in the ~1-month window the founder already identified.

For the phData pivot, this sharpens the open action rather than changing it. The career-fork doc's instinct to "lock the explicit first-pull commitment with Andrew" and "bank the CCA signal early" is correct, but the reason is more concrete than "Tiger-readiness": on the adoption path the founder personally supplies all three Services Track ingredients at once — a Claude-certified architect (cert leg), working named deployments like Lionsgate (production-customer leg), and, if he bakes a customer-story permission clause into engagements, the public-story leg. That triple is the only currency Anthropic's model actually counts, and it is also the EoY negotiating artifact inside phData. The cert is genuinely table stakes on this route, which strengthens the existing "sit CCA before Nov 22" lean — but note the timing nuance: the cert helps phData's firm counter only if his exam registration is linked to phData's partner record (open ask #5 in the prior brief), so the founder should confirm that linkage when he locks the first-pull conversation, not after.

For RDCO's own Anthropic positioning, the read is sobering and clarifying. The Services Track tiers are headcount-gated by design — Select alone needs 10 active-certified individuals — so a solo-founder holding company cannot meaningfully tier on the firm ladder, and chasing "partner status" for RDCO directly is a non-starter. RDCO's Anthropic leverage is not a tier badge; it's the founder being an individually-certified, deployment-proven Claude architect whose credential and reference-able work accrue to whatever firm he's attached to (today phData). The portable asset is the individual evidence ledger, not an RDCO partner listing. If RDCO ever wants direct Anthropic ecosystem presence, the realistic surface is the builder/marketplace side (skills, MCP connectors, published IP) rather than the SI Services Track — worth a separate scoped look, not assumed.

Calibration: tier thresholds and the absence of a "Tiger Team" program are well-sourced (Anthropic primary + independent analyst + multiple press). The named-firm list is solid for the GSIs and TCS/DXC; the EPAM detail is search-level only (its press release 403'd, not fetched). Andrew's intended requirements for his team are genuinely unknown to us and should not be inferred.

Open follow-ups

Related

Sources

Vault:

Web (fetched this run):

Web (search-level / not fully fetched):