Kautilya PE · Full-cycle acquisition advisory · AI
Three serious deals.Two walked away from.One closed for $12,000.
A first-time buyer with a technical background, capital, and no defined mandate. Twelve weeks later: a written mandate, a five-gate screen, two deals killed (one after the LOI), and an all-cash close on a GPT education platform with 2.5M conversations.
- Client
- First-time acquirer, technical background
- Asset
- SmartPrompt · 30 to 35 GPTs
- Deal size
- $12,000, all cash
- Timeline
- 12 weeks · Jul to Sep 2024
- Cumulative conversations
- 2.5M
- Conversations a month
- ~100K
- Targets identified
- ~300
- Earn-outs or contingencies
- 0
The situation
Capital, intent, and no decision framework
The client came in with engineering expertise, capital to deploy and a clear preference for buying over building, specifically in AI and SaaS. What was missing was everything that turns intent into a good acquisition: no sector focus, no business model preference, no stated risk tolerance, and a budget that was flexible but not matched to any category.
The core problem wasn't a shortage of deals. It was decision architecture. Evaluation leaned on how appealing a sector felt rather than how the business ran, several unrelated opportunities were live at once, and nothing in the process would stop a premature commitment.
What could go wrong without a framework
- 01Deploying capital into an operationally complex asset beyond the buyer's current skill set.
- 02Accepting bad terms out of enthusiasm for the sector.
- 03Missing structural risks in the first assessment.
- 04Executing on a deal with asymmetric downside.
The asset
A GPT portfolio with something most GPTs don't have
SmartPrompt surfaced in week 9 and immediately looked different from the rest of the deal flow. Its edge is user-level usage visibility, which OpenAI does not provide for standard GPTs. That visibility is what makes targeted monetisation, data-driven product work and measurable partnerships possible.
- Portfolio
- 30 to 35 GPTs, niche education focus
- Cumulative conversations
- 2.5M
- Monthly conversations
- ~100K
- Active users
- ~10,000
- Operating history
- ~2 years
- Revenue
- Peak MRR ~$300; minimal at acquisition
- Distribution
- Embedded through the GPT Store
- Overhead
- Low; modest technical management
Sourcing
300 targets, three serious candidates
Off-market outreach through LinkedIn, Twitter/X, niche Reddit communities and direct founder contact, aiming for three to four qualified opportunities a week.
Tap a stage for what happened there.
1 · Closed
SmartPrompt: $12,000 all cash, APA signed within 7 days of the LOI.
The approach
Mandate, framework, sourcing, execution
In that order, because each step depends on the one before it.
- 01
Mandate architecture · weeks 1 to 2
Deliberate exposure to deal flow across content and education, real-estate-adjacent businesses, AI tools and small founder-led products, watching which ones the client leaned into and why.
- 02
A five-gate evaluation framework
Financial reality, risk identification, market validation, price discipline and deal structure. Every opportunity had to clear each gate in sequence; failing one meant disqualification, and sector enthusiasm alone never moved a deal forward.
- 03
Off-market sourcing · weeks 3 to 9
A continuously refreshed pipeline, with criteria tightened in weeks 5 to 6 toward AI-native, GPT-driven platforms with usage visibility and operational independence.
- 04
Diligence, close and handover · weeks 10 to 12
Portfolio audit, ownership validation, asset capture and platform-dependency review. An operator was recruited in parallel so the business had someone running it the week it closed.
Mandate
From open-ended to specific
Two weeks of structured exposure turned an exploratory brief into this.
Kill discipline
Three candidates, one framework
Each serious candidate ran the same five gates. Pick one to see where it passed, where it broke, and what we advised.
30 to 35 GPTs, 2.5M conversations, ~10,000 active users, usage visibility.
- 1Financial realityPass
- 2Risk identificationPass
- 3Market validationPass
- 4Price disciplinePass
- 5Deal structurePass
Decision
Closed at $12,000, all cash
Clean cash flow, low operational complexity, acceptable platform dependency on ChatGPT, and a price anchored on present risk rather than platform upside. No earn-outs, no contingencies.
What diligence revealed
Why this one cleared every gate
Usage visibility is the moat
What we found
A 30 to 35 GPT portfolio with variable but acceptable quality, a substantial reusable knowledge base, and user-level engagement data that standard GPT assets don't expose.
What it meant for the buyer
The buyer could monetise, prioritise and partner using real usage data instead of guesses, with management requirements that matched a technical owner's skill set.
The result
A clean close with nothing left hanging
LOI in week 9, diligence done in week 10, APA signed and assets transferred in week 11. $12,000 bought all 30 to 35 GPTs, the full knowledge base, the GPT Store account, domain and brand assets, the LinkedIn presence and all intellectual property.
Just as important, the client came out with a replicable playbook: a clear mandate, a gate-based evaluation process, and the proven willingness to walk away from a deal after the LOI.
- All cash, 100% equity
- $12,000
- LOI to signed APA
- 7 days
- Deals killed on the way
- 2
Outcome
The right acquisition is worth the wait. The wrong one, whatever the sector appeal or deal momentum, is always worth walking away from.
Deliverables
What the client received
Acquisition mandate
An exploratory brief turned into a defined sector, model, risk profile and scale.
Five-gate framework
A sequential screen applied to every opportunity, with disqualification on failure.
Off-market sourcing
~300 targets, ~50 founder conversations, three serious candidates.
Kill analysis
A pre-LOI withdrawal on ResearchKick and a post-LOI termination on Keymate AI, both documented.
Technical diligence
GPT portfolio audit, ownership validation, asset capture and platform dependency assessment.
Close and transition
Negotiation, APA, operator recruitment and a three-phase handover to full client control.
Somebody is selling your next project.
They just haven't met you yet. Next issue goes out Tuesday, 7am ET.
Free. Two emails a month. One-click unsubscribe.