Helping PE Sponsors and CEOs Manufacture the Return: Underwrite to Exit, and Everything in Between
Value creation run as one operating system - underwriting, the 100-day plan, pricing, cash, operations, organization and the hold decision - proven in cash, not claimed on a slide. AI is one mechanism among many, used where it deepens returns.
The Value-Creation Chain·Seven stages·Twelve levers·Hover to preview, click to read·Tap to read
CHAIN · 01
What must be true, and how much to pay for.
CHAIN · 02
Three to six bets, owned by name.
CHAIN · 03
Capability, capacity, cadence, incentives.
CHAIN · 04
Identified, contracted, realized.
CHAIN · 05
Would we buy it again, at today’s value?
CHAIN · 06
Make portco #2 cheaper than #1.
CHAIN · 07
Prove the story rather than invent it.
CHAIN · 01
What must be true, and how much to pay for.
CHAIN · 02
Three to six bets, owned by name.
CHAIN · 03
Capability, capacity, cadence, incentives.
CHAIN · 04
Identified, contracted, realized.
CHAIN · 05
Would we buy it again, at today’s value?
CHAIN · 06
Make portco #2 cheaper than #1.
CHAIN · 07
Prove the story rather than invent it.
LEVER · A
Only realized price counts.
LEVER · B
Where growth will actually come from.
LEVER · C
Negotiated is not realized.
LEVER · D
Release cash already earned.
LEVER · E
When the constraint is physical.
LEVER · F
Integrate, don’t aggregate.
LEVER · G
Role by role, against value at stake.
LEVER · H
What people are really paid to do.
LEVER · I
Hours are not EBITDA until converted.
LEVER · J
Altitude, five conversations, one question.
LEVER · K
Hold, sell, recapitalize or continue.
LEVER · L
Make the value-creation claim believable.
Independent Operating Advisor · Boston · Global
The return now has to be manufactured inside the portfolio company and proven in cash. I advise a small number of PE-backed, mid-market and upper mid-market enterprises each year - directly, as a founding COO who scaled one from inception to a ~$1 billion outcome.
Why Now
For a decade, cheap leverage and rising multiples did much of the work: between 2010 and 2022 they delivered 59% of buyout returns, against 41% from revenue growth and margin (McKinsey). That tailwind has gone. What remains is the operating work - and a far higher bar for evidence from LPs, buyers and boards.
Annual EBITDA growth a buyout now needs to earn 2.5x in five years - against roughly 5% for the same result in 2015.
Bain & Company · Global Private Equity Report 2026
Distributions as a share of private-equity AUM in the year to June 2025, against a 16% average in 2015-19.
McKinsey · Global Private Markets Report 2026
Of sponsors realized less than three-quarters of the value their value-creation plans called for.
Alvarez & Marsal · Value Creation Survey 2025
Of companies attribute any EBIT impact to AI - even though nearly nine in ten now use it.
McKinsey · The State of AI 2026
The Practice
Most value-creation advice is written by people who have never carried a P&L, sat in a deal review, or lived with a plan after the deal team moved on. I have - for thirteen years - as founding COO of a PE-backed enterprise that scaled from inception to $280M+ revenue, ~$1B enterprise value, 40%+ EBITDA margins, and an 8-10x MOIC for early sponsors.
I now point that operating lens at the question every sponsor is being asked by its LPs: how is the return actually being made - underwritten at entry, delivered through the hold, and proven at exit - now that leverage and multiple expansion can no longer carry it?
The answer is rarely one lever. It is price realization and the commercial engine, cash, operations, organization and incentives, buy-and-build and capital allocation - sequenced to what management can carry. AI is increasingly one of the mechanisms beneath those levers, and I treat it the way a sponsor treats any capital: fund it where it earns its cost, stop it where it does not.
I do not sell licenses. I do not staff projects. I do not write slides on top of your team's slides. I sit at the operating elbow of the CEO - and, where relevant, the sponsor's operating partner and the chair - and raise the quality of the decisions that determine the exit.
“You want an advisor who has already done the job you are asking your CEO to do - in the geometry and timebox a sponsor expects.”
The Value-Creation Chain
Seven stages, each with the question it must answer, the artifact it must leave behind, and the failure it exists to prevent - and twelve levers that run along the chain. Every stage and lever carries a scored worksheet. The scores stay locked until an engagement begins; the thinking behind them is open. Select any one to read it.
Where the value actually comes from. AI sits beneath them as a mechanism - Lever I - not above them as a thesis.
Scorecards shown are illustrative. Scored worksheets travel under engagement.
Request a value-creation reviewApproach
Each discipline carries its own cadence, its own written artifacts, and its own line to the board. They are not services. They are the joints of a single operating system for value creation - from the investment committee to the closing dinner.
A written value-creation thesis - not a workshop deck - that separates what is executable from what is merely possible, prices only the first, and concentrates scarce management attention on the three to six bets that move enterprise value.
The levers that still create most of the value - price realization, the revenue system, procurement, working capital and core operations - each tracked from identified to contracted to realized, and signed off by Finance.
Structure, talent, incentives and cadence built around the plan - with AI deployed as an execution mechanism beneath the levers and governed like capital: fund, prove, scale or stop.
Board time spent on decisions rather than reporting, a scheduled re-underwrite that actually reallocates capital and people, and an equity story proven over six quarters rather than invented six months out. Built by someone who has lived through a $500M recapitalization, a $540M partial exit and a signed $1.2B transaction.
Engagements
Every engagement is bespoke. These are the shapes it usually takes. Fees are commensurate with scope, seniority and the capital at stake; they are discussed privately once fit is established.
Sponsor-side
Sit inside a PE firm's value-creation team for a portfolio cohort. Operating diligence on new deals, 100-day plans for platform and add-on investments, re-underwriting reviews across the book, and direct CEO coaching.
Portfolio-side
Named advisor to a portfolio company CEO through a defined value-creation window - a pricing or commercial reset, a hold-period transformation, a carve-out, a bolt-on integration, or the run-up to a sponsor transaction.
Board-side
A single board or committee engages me as independent counsel on value-creation reviews, capital allocation, re-underwriting and exit readiness - between board meetings, before critical decisions, and on AI strategy and risk where it is material.
Sponsor / LP Education
A 100-hour curriculum, underwrite to exit, for a single firm's partners, operating team, or a portfolio-company C-suite - with AI taught as a mechanism of value creation, from atoms to strategy. Privately delivered. Written from the operating chair, not the classroom.
Track Record
Saurabh is an independent operating advisor based in Boston. He spent thirteen years as founding COO of ACT Fibernet, India's leading fiber-optic ISP, scaling the business from inception to ₹22 billion in revenue (about $280M), ~2M broadband and ~1M digital-services customers, and unicorn enterprise value. He served as the principal sponsor-facing operating interface through a $500M recapitalization, a $540M partial exit to a large US private-equity house, and a subsequent signed $1.2B transaction.
Earlier, at Procter & Gamble, he held a series of leadership roles, including running one of the company's fastest-growing regional P&Ls. He began his career as an engineer with the Tata Group.
He holds an MBA from the MIT Sloan School of Management, a PGDM from the Indian Institute of Management Calcutta (full scholarship), and a B.Tech (Hons) in Electrical Engineering from the Indian Institute of Technology Kharagpur.
He was featured as a founding COO in Credit Suisse's 2021 100 Unicorns report and received the Outstanding Leadership Award at the Internet 2.0 Conference (2023).
The AI Workbooks · Beneath Lever I
AI is one of the mechanisms beneath the value-creation levers - increasingly an important one. When it is the right one, these twenty-five board-grade worksheets underwrite the bet, tier the risk and hold the deployment to Monday-morning reality, until the hours it releases show up as EBITDA or cash. A glimpse of the library; the workbooks themselves travel under engagement.
Two AI workbooks·Twenty-five frameworks·Assess, then stack
ASSESS · 01
Eight dimensions, scored zero to five.
ASSESS · 02
Where to play, how to win, what to stop.
ASSESS · 03
70 / 20 / 10 across three horizons.
ASSESS · 04
The screen before the business case.
ASSESS · 05
Which layer to build, which to buy.
ASSESS · 06
Board-grade: approve, pilot, defer, reject.
ASSESS · 07
Six dimensions, four kinds of risk.
ASSESS · 08
EU AI Act, NIST, the director's questions.
ASSESS · 09
Automate, augment, elevate, eliminate.
ASSESS · 10
What becomes of the hours you free.
ASSESS · 11
The decisions that earn the data.
ASSESS · 12
The six-layer enterprise stack, stress-tested.
ASSESS · 13
The leader's own operating system.
ASSESS · 14
The mirror most boards skip.
ASSESS · 15
Thirty, sixty, ninety - and the north star.
STACK · A
Where AI value is underwritten, not asserted.
STACK · B
Three deployment lanes, one discipline.
STACK · C
Weighted model and platform choice.
STACK · D
Low to restricted - and the controls each demands.
STACK · E
The harness every production use case needs.
STACK · F
Access is not adoption.
STACK · G
Scale only on sustained movement.
STACK · H
Where accountability actually sits.
STACK · I
The human half of the transformation.
STACK · J
FDE, consultancy, and the internal team.
Perspectives
A small set of long-form essays and briefings for boards, sponsors and CEOs - on value creation first, and on AI where it is one of the mechanisms. More is published privately to engaged clients than appears here.
Why leverage and multiple can no longer carry the return, and what replaces them: the executability test at underwriting, the three-to-six-bet value bridge, the realization ledger, the scheduled re-underwrite, the board's five conversations, and the six-quarter evidence trail a buyer will pay for.
Request the briefing →Only about a third of sponsors run a structured re-underwrite during the hold. The question worth asking every year - would we buy this company again, at today's value, with what ownership has taught us? - and what to do with the answer.
Request the essay →Nearly nine in ten companies use AI; little more than a third can point to any EBIT impact. The four routes by which released capacity becomes earnings or cash - and the board question that separates a productivity story from a value-creation one.
Request the essay →A director's checklist - inventory, accountable owner, EU AI Act exposure, vendor data terms, incident response, bias regime, model lifecycle. Written in the language the board already speaks.
Request the essay →The prompt → retrieve → fine-tune → train ladder, read from the P&L rather than the paper. Where the moat is, where the vendor risk is, and what a sponsor should underwrite at this valuation.
Request the essay →In professional services and in service-heavy portfolio companies, AI collapses the bottom of the pyramid. The firms that win will rebuild the learning model deliberately. A note to CEOs and sponsors who still staff the old way.
Request the essay →Engage
I take on a limited number of engagements at any one time. If a sponsor, a chair or a CEO is weighing a decision that moves enterprise value - an underwriting call, a 100-day plan, a pricing or commercial reset, a re-underwrite, a carve-out, a transaction, or an AI bet that has to earn its cost - I'm happy to have a private conversation and tell you, honestly, whether I am the right advisor or whether you need someone else.
Prefer email? saurabh@saurabhmukherjee.com