OpenAI just launched a $4B 'consulting company' — and lined up TPG, Brookfield, SoftBank, and Goldman behind it

The first time a model maker built its own consulting firm — May 11, OpenAI Deployment Company

Here's the deal. On May 11, OpenAI officially launched the OpenAI Deployment Company (ODC). The capital base alone is over $4B. TPG leads, with Advent, Bain Capital, and Brookfield as co-lead founding partners. SoftBank, Goldman Sachs, Warburg Pincus, B Capital, BBVA, Emergence Capital, Goanna, and WCAS round out the founding cohort. Nineteen global investment, consulting, and SI partners in total. Bloomberg pegged the post-money valuation at $14B ($10B pre-money) right after the announcement. Ownership: OpenAI majority and control.

The second card landed simultaneously — OpenAI is acquiring UK-based AI consultancy Tomoro. Tomoro runs a global top-tier client base — Tesco, Virgin Atlantic, Supercell, Mattel, Red Bull — and specializes in enterprise GenAI deployment. About 150 Forward Deployed Engineers (FDEs) and Deployment Specialists join ODC from Day 1. So ODC starts life with revenue, top-tier references, and a staffed talent pool already in place.

Why this isn't just another consulting subsidiary launch comes down to timing. Exactly one week earlier, on May 4, Anthropic announced a $1.5B PE joint venture with Blackstone and Goldman Sachs. Two frontier labs in the same cycle made the same bet — "the model company comes downstream all the way to enterprise deployment." OpenAI CRO Denise Dresser told CNBC that "enterprise AI adoption is at a tipping point," citing OpenAI's own survey: 97% of executives have deployed AI agents in the past year, but 79% of organizations face challenges adopting AI. The gap between those two numbers is the ground ODC is moving into.

Each player — OpenAI, ODC, Tomoro, the 19 partners

OpenAI. With this launch OpenAI has now planted a flag in models + infrastructure (Stargate) + consumer (ChatGPT) + enterprise consulting (ODC) — all four axes. May ARR is estimated in the high $20Bs, with enterprise share growing fastest. ODC promotes that enterprise line from "API sales" to "full-stack deployment partner." CEO Sam Altman didn't comment publicly, but per industry reporting the internal vision is "every company that uses ChatGPT will eventually have OpenAI engineers in their building."

OpenAI Deployment Company (ODC). Legally a majority-owned and controlled OpenAI subsidiary. Post-money valuation ~$14B ($10B pre-money), $4B+ in new capital. Brookfield committed $500M alone via Brookfield Business Corporation. ODC runs two revenue lines simultaneously — (1) consulting revenue from Tomoro's existing client base, (2) net-new revenue when OpenAI API customers need deployment help and route to ODC. The margin structure is closer to "AI deployment outcome" pricing than McKinsey's billable-hour model.

Tomoro. Founded 2023 in the UK, focused entirely on GenAI enterprise deployment. ~150 employees at acquisition, mostly FDEs or Deployment Specialists. The client base is interesting — Tesco (UK's largest grocery chain), Virgin Atlantic (airline), Supercell (gaming), Mattel (toys), Red Bull (consumer). B2C, retail, and media are all in the mix, suggesting OpenAI deliberately bought a "frontier model → traditional industry" distribution channel. Tomoro brand gets absorbed into ODC.

The 19 partners. TPG leads. Advent, Bain Capital, Brookfield co-lead. Founding partners also include B Capital, BBVA, Emergence Capital, Goanna, Goldman Sachs, SoftBank, Warburg Pincus, WCAS — plus seven additional consulting/SI partners. What this configuration delivers: (1) PE giants bring their portfolio companies (hundreds of private firms) as ODC's first client pool, (2) BBVA, Goldman, and SoftBank open financial-services beachheads, (3) Bain and Advent contribute consulting know-how and talent training, (4) Brookfield provides infrastructure funding and a global real-estate/energy deployment channel. This isn't capital — it's a market-penetration package.

Denise Dresser (OpenAI CRO). Eighteen years at SAP, eleven at Salesforce — career enterprise sales operator. After joining OpenAI in 2024, revenue reportedly grew 4x in her first year. The day-to-day operator behind ODC.

The substance — capital structure, valuation, FDE model

Capital structure. $4B+ in new equity into ODC. Brookfield alone wrote a $500M check via Brookfield Business Corporation. The rest distributes across TPG, Advent, Bain, SoftBank, Goldman, and others. Valuation marks at $14B post-money, $10B pre-money — meaning $4B of new capital takes the company to a $14B mark immediately. Translation: ODC is the second most valuable OpenAI-controlled asset on day one. The first is OpenAI itself (~$500B).

Item	Value
Total new capital	$4B+
Valuation (post-money)	$14B
Brookfield commitment alone	$500M
Founding partner count	19
Day 1 FDE headcount	~150
Tomoro acquisition stage	Full UK consultancy absorbed

Forward Deployed Engineer model. FDE is the operating heart of ODC. The concept comes from Palantir — unlike a generalist consultant, an FDE is (1) a full-stack engineer, (2) embedded onsite at the client for 6–18 months, (3) re-architecting model + data + workflow as a single bundle. A McKinsey senior consultant bills $1,500–$2,500/hour for slides; an FDE costs about the same but ships production code, deployed models, and a trained client team. By absorbing Tomoro's 150 staff, ODC turns this model on at global scale on day one.

Cost and pricing. Per Bloomberg and Axios reporting, a standard ODC engagement runs $5–25M over 6–12 months; full enterprise rollout runs $50–200M over three years. McKinsey's average digital-transformation engagement is roughly $30–80M over three years, so ODC prices broadly in line — slightly higher in some bands. The differentiators McKinsey can't structurally match: (1) FDEs have direct access to OpenAI model internals, (2) priority access to new model releases, (3) priority allocation of Stargate compute. Together these form a real moat.

Competitive landscape. McKinsey, Accenture, Bain, BCG, and Deloitte are each building "AI deployment" arms — collectively over $150B in AI consulting revenue annually. ODC isn't taking a slice of that pool; it locks API and consulting revenue together, so the impact is bigger than the sum of the parts. Anthropic's May 4 Blackstone/Goldman JV ($1.5B) is the counter, but it's less than half the size in capital and partner count. Google and Microsoft have internal consulting arms (Google Cloud Consulting, Microsoft Industry Solutions Delivery), but neither has built the same independent-capital + PE-portfolio package.

Who gets what

OpenAI's win. First, revenue diversification. API + ChatGPT Pro + Enterprise + ODC is now a four-line business. Second, data channel. FDEs get into client workflows and (within contractual scope) become a legal pipeline for domain training data. Third, lock-in. Once ODC is inside a company, switching off OpenAI is functionally impossible — full-stack deployment is welded to OpenAI model assumptions.

TPG / Brookfield / Bain Capital / Advent's win. For PE, ODC equity is (1) likely to mark up from $14B to $40–60B in 18–24 months (Anthropic's JV was 6x smaller at launch), (2) a portfolio prioritization right (deploy ODC into our LBO targets first), (3) a flagship asset to anchor a follow-on "AI Transformation Fund." Brookfield's $500M check is already marketing as the first deployment of an "AI Infrastructure Fund."

SoftBank / Goldman / BBVA's win. SoftBank now has full-stack OpenAI exposure across models, infra, and services. Goldman, already running OpenAI in trading desks, just extended that relationship to the cap table. BBVA gets early access on European and Latin American banking deployment bids.

McKinsey / Accenture's loss. AI transformation is the 70–100% YoY growth engine inside both firms. Once ODC is operating, (1) the model owner shows up directly and pressures price, (2) clients start asking "why route through a consultancy at all?" Expect AI-transformation revenue growth to slow at both firms over the next 18 months.

Anthropic's loss. The May 4 Blackstone/Goldman JV ($1.5B) just got dwarfed in the same news cycle. If $14B becomes the comparable, Anthropic faces fundraising pressure on the next round. Anthropic still has its narrower financial-services focus as a defensible niche.

Past parallels — wins and losses

Win: IBM Global Services (1990s–2000s). IBM spun up IGS in 1991 and bought PwC Consulting in 2002 for $3.5B. Result: services became more than half of IBM revenue and carried the company through hardware decline. ODC's model is closest to this. Difference: IBM was hardware-going-services; OpenAI is models-going-services. The model business has structurally better margins, so the transition could be even more profitable.

Win: Palantir Foundry + FDE model. Since 2003 Palantir has run FDEs delivering full-stack deployments to government and industry. Result: ~$150B market cap and ~$3B revenue by 2024. The OpenAI-Tomoro deal is unmistakably a Palantir copy. Reports suggest some Palantir alumni are joining ODC.

Loss: HP–EDS (2008, $13.9B). HP bought EDS to push into IT services; six years later it took an $8B impairment. Why it failed: (1) parent product and service revenue cannibalized each other, (2) culture and incentive mismatches between service and product staff. ODC could replicate this — if Tomoro's 150 staff misalign with OpenAI's incentive structure and 50% leave in 18–24 months.

Loss: SAP–Sybase (2010, $5.8B). SAP bought a database company and tried to integrate it; eventually run as a separate subsidiary and divested. The integration failed because the two firms had different client bases. Tomoro (retail/CPG-heavy) and OpenAI Enterprise (tech/finance-heavy) face the same risk. Whether they actually fuse is the central question of the next 12 months.

Competitor counter-plays

Anthropic. Most likely fast follow. Rather than scaling Blackstone/Goldman by raw capital, expect deeper vertical specialization. The May 5 Claude Opus 4.7 + 10 financial-services agents launch is the signal. ODC is horizontal (every industry); Anthropic can win by going deeper in finance, legal, healthcare. But within 30 months Anthropic will face pressure to expand industries.

Google DeepMind. Strengthen Google Cloud Consulting and bundle Gemini Enterprise + Workspace + Vertex AI as a full-stack package. Google's weakness is sales — neither as deep nor as motion-fast as OpenAI or Microsoft — making it harder to assemble an ODC-style PE + consulting integration.

Microsoft. Holds the largest counter-card. Microsoft Consulting Services already has 50,000 staff, and Azure + Copilot deployment is in motion — direct overlap with ODC's hunting ground. But Microsoft owns 49% of OpenAI, so the rational path is using ODC as a channel to grow Azure rather than competing head-on.

McKinsey / Accenture / Deloitte. Short term: (1) carve out AI-deployment business units as JVs or spin-outs, (2) deepen partnerships with non-OpenAI labs (Anthropic, Google, xAI), (3) lock clients into multi-year retainers. Medium term (12–24 months): expect margin compression on AI-deployment.

Salesforce / SAP / Oracle / ServiceNow. Enterprise SaaS giants. They're deploying their own AI agents (Salesforce Einstein, SAP Joule). The defining question for ODC's first 12 months: does ODC sit on top of these stacks as a deployment partner, or compete with them?

Cognizant / Infosys / TCS / HCL. India-based IT services giants. Their cost advantage lets them position as "low-cost AI deployment" and capture the mid-market. But without direct model-maker access, they'll struggle to differentiate at the high end.

So what changes — by persona

CTO / CIO. If your stack today is "OpenAI API + in-house engineers or a consultancy," "ODC full package" is now a third option. Decision criteria: (1) deployment speed (ODC may be 6–12 months faster than building it yourself), (2) lock-in cost (post-ODC, swapping models is effectively impossible), (3) unit cost (can you absorb a $5–25M starting engagement?). ODC isn't automatically the right call — it makes most sense in domains where AI urgency is high (customer service, back-office automation).

Startup founder. Two effects. First, AI-agent SaaS startups now face ODC as a direct competitor in the enterprise deployment market. Second, on the flip side, ODC engagements that don't go well will spawn "rescue work" for startups with the right specialization. The likely segmentation: mid-market and SMB go to startups, F500 and F1000 go to ODC.

Investor. IPO windows narrow for AI consulting and SI players. Multiples compress as the model maker enters directly. Vertical AI deployment specialists (legal, healthcare, manufacturing) — areas ODC won't go directly — keep their multiples. On the PE side, ODC becomes the template; expect "AI deployment company" rounds at other model labs to follow.

Developer / engineer. FDE becomes a new tier-one career path. The Palantir-born model is now standard at OpenAI, Anthropic, and Google — expect "AI deployment engineer" salaries to run 30–50% premium to senior SWE comp. Trade-off: heavy onsite client work means a lifestyle hit.

Office worker. If your employer adopts ODC, expect 30–50% of your workflow to be agent-automated within 12–24 months. ODC's value prop is "deploy fast," so transformation will run 2–3x faster than typical "AI transformation" projects. Pivoting your role to meta-work — managing and evaluating agents — is the rational move.

Regulator. ODC raises three new regulatory issues — (1) data privacy (FDEs touch client internals), (2) competition law (model maker vertically integrating into consulting), (3) AI safety (who owns deployment validation responsibility). How EU AI Act maps onto ODC will be a defining question in the next 6–12 months.

References

• OpenAI: OpenAI launches the OpenAI Deployment Company to help businesses build around intelligence
• Bloomberg: OpenAI to Buy Consulting Firm for Private Equity Joint Venture
• Axios: OpenAI launches AI consulting arm valued at $14 billion
• CNBC: OpenAI revenue chief Dresser says enterprise AI adoption is 'at a tipping point'
• Tomoro: Tomoro Acquired By OpenAI Deployment Company
• Bain & Company: Bain & Company invests in the OpenAI Deployment Company
• BBVA: BBVA joins OpenAI's new company to accelerate AI enterprise transformation