No PriorsAmex Global Business Travel: The World’s First AI Take Private with Long Lake CEO Alexander Taubman
CHAPTERS
Alexander Taubman and Long Lake’s “AI take-private” thesis
Elad Gil introduces Alexander Taubman, CEO of Long Lake Management, framing the $6.3B intent to acquire Amex Global Business Travel as a first-of-its-kind “AI take-private.” The conversation sets up Long Lake’s broader model: acquiring services businesses and improving performance through applied AI.
- •Long Lake announces intent to acquire Amex Global Business Travel for $6.3B
- •Positioned as the “world’s first AI take-private” deal
- •Long Lake has completed ~30 prior acquisitions using an AI-transformation playbook
- •Core concept: buy real-world services businesses and optimize them with AI
Inside Nexus: Long Lake’s horizontal AI platform and deployment process
Taubman explains Nexus, Long Lake’s shared AI infrastructure used across multiple verticals. He emphasizes that the hard part isn’t model selection but workflow mapping, data cleanup, integrations, and applied engineering that connects models to real operational work.
- •Nexus is a horizontal platform: ~80% shared infrastructure across verticals
- •Long Lake is model-agnostic; Nexus sits between models and business workflows/data
- •Deployment involves workflow mapping, identifying/cleaning data sources, and integration work
- •Early rollouts took >1 year; now deployment can show impact within days
AI as a growth engine, not a cost-cutting program
Rather than focusing on layoffs or pure margin expansion, Long Lake uses AI to create time savings that are reinvested into better customer service and faster growth. Taubman describes turning slow-growth services businesses into “software-like” growth profiles by lowering incremental costs while improving experience.
- •AI creates immediate time savings; Long Lake reinvests that capacity into growth
- •Focus areas: customer experience, responsiveness, and service quality
- •HOA portfolio example: shifting from ~0–5% growth to 20%+ organic growth
- •Improved unit economics: serve more customers at lower incremental cost with better outcomes
Retention and the talent flywheel from “AI superpowers”
Taubman argues that once employees experience reduced busywork and better tooling, they are reluctant to move to competitors that lack automation. Higher productivity enables higher compensation, which attracts better talent, which further improves customer outcomes—reinforcing a self-perpetuating flywheel.
- •High employee retention across acquisitions due to reduced mundane work
- •Leaving Long Lake means returning to manual tasks (25–30% of the day)
- •Higher productivity supports higher pay, improving recruiting competitiveness
- •Customer outcomes improve: faster responses, fewer errors, higher customer retention
Why acquire businesses instead of selling AI software to them
The discussion contrasts the Silicon Valley vendor approach with Long Lake’s ownership model. Taubman emphasizes alignment: owning the business ties AI implementation directly to business outcomes and enables the change management required for real adoption.
- •Acquisitions create deeper alignment than vendor/customer relationships
- •Owning customer relationships enables end-to-end accountability for outcomes
- •Internal product customer is the frontline employee, creating a tight feedback loop
- •Change management is easier when Long Lake controls processes and org design
Building the founding team: combining M&A, engineering, and change management
Elad highlights that AI roll-ups require three rare competencies: dealmaking, technical execution, and operational change. Taubman explains Long Lake was purpose-built to integrate these skill sets, initially hiring heavily through trusted networks, including engineers from top tech companies and founders of applied AI startups.
- •Long Lake designed from day one to blend PE/M&A, engineering, and change management
- •Early hires (first ~20) were network-based for trust and speed
- •Engineering talent drawn from companies like Palantir, Ramp, Robinhood, Glean
- •Many technical hires had founder backgrounds and wanted to bring AI into real-world services
M&A bench strength: attracting elite private equity talent to an AI-native platform
Taubman outlines how Long Lake’s investment team is staffed by professionals from major PE firms. The differentiator is an AI-native operating model: M&A professionals join for the chance to execute deals with a technology advantage, not just financial engineering.
- •M&A/PE talent sourced from firms such as GTCR, Blackstone, TPG, HIG
- •Long Lake appeals to investors who believe traditional PE is not AI-native
- •The platform thesis: operational outperformance through applied AI, not leverage alone
- •Creates a distinctive environment for deal professionals seeking AI-driven transformation
The Amex GBT take-private: why corporate travel fits Long Lake’s “prepared mind” list
Taubman shares the strategic rationale (within public-deal constraints): travel was a pre-identified target industry due to mission-criticality and high cost of failure. He emphasizes the franchise strength and trust built over a century and frames the acquisition as an opportunity to accelerate an already-existing AI roadmap.
- •Amex GBT is ~111 years old (roots in WWI-era American Express traveler support)
- •Travel viewed as mission-critical, high-trust, high-cost-to-failure service
- •Long Lake maintained a “whiteboard” list of priority industries; travel was on it
- •Intent is to double down on Amex GBT’s existing AI transformation strategy
AI-enabled travel service: the “counselor with superpowers” vision
Taubman describes the practical product vision: augmenting travel counselors so they can resolve disruptions faster and deliver better experiences. The emphasis remains on empowering humans with AI, rather than replacing the service model entirely.
- •Goal: faster response times and better disruption resolution for travelers
- •Conceptual product: a travel counselor augmented by AI “superpowers”
- •Nexus model: use AI to enhance frontline service delivery
- •Customer excellence positioned as the central outcome of AI deployment
Long-term ownership model: Berkshire/Danaher-style compounding over quick flips
Elad contrasts Long Lake with traditional short-term private equity. Taubman explains why multi-year operational transformation and flywheel effects require patient capital—once a category-leading services business is built, Long Lake aims to hold and compound rather than sell.
- •Rejects the short-term PE playbook of quick flips and aggressive cuts
- •Transformation is multi-year; benefits compound through talent and customer flywheels
- •Inspired by long-term compounders like Danaher (and Berkshire-like stewardship)
- •Vision: be a long-term owner and steward for employees, customers, and founders
Winning deals: permanent capital plus day-one AI capability and aligned rollover equity
Taubman explains why sellers choose Long Lake even in competitive processes: the offer combines long-term partnership with immediate operational help from embedded engineers. Long Lake often encourages founders/management to roll equity, aligning incentives so everyone benefits from AI-driven productivity gains.
- •Value proposition: long-term partner with deep applied AI engineering resources
- •AI underpenetration (~1%) creates strong demand for credible implementation help
- •Engineers “live in your office” to drive change management and execution
- •Founder/management rollover equity is encouraged to share upside and align incentives
Making services companies feel like software businesses: scaling growth with higher incremental margins
The episode closes by detailing why AI changes the growth math for labor-intensive services. When teams become 30–40% more efficient, companies can grow without hiring proportionally, improving incremental margins and making growth enjoyable again for long-time operators.
- •Traditional services growth requires proportional hiring, training, and management
- •High labor share creates low incremental profit on new revenue
- •AI-driven efficiency (30–40%) allows growth without matching headcount increases
- •Result: “software-like” scaling—more investment capacity, happier teams, happier customers
