Manlio Carrelli shares analysis of AI companies, enterprise technology, and market shifts, including how businesses can use predictive intelligence to identify potential customers, partners, and acquisition targets.
Manlio Carrelli is most useful for data-backed analysis of AI company momentum, enterprise M&A, and market shifts.
Best for
A B2B AI market intelligence firm wanting a co-authored report on AI startup momentum and enterprise M&A trends.
Not for
A developer-tools vendor wanting hands-on product tutorials, recorded walkthroughs, live demos, or paid-media reach guarantees.
Key takeaways
One post ranks eight AI companies whose success-probability score rose 300 points or more.
Coinbase's investments, partnerships, acquisitions, and hiring across five business areas suggest an everything exchange.
Public software incumbents' purchases of AI startups are projected to more than double year over year.
Check first: Ask for audience analytics, budget, turnaround, editorial control, approval process, exclusivity, and disclosure terms.
Main ideas
Predictive intelligence beats waiting for the news · Incumbents are buying the disruptors, not being replaced by them · The budget AI actually replaces is labor, not software
Manlio Carrelli explains how AI is changing companies, markets, and work.
In their LinkedIn posts, they track signals such as startup performance, hiring, acquisitions, and partnerships to show where a market may be heading before a shift becomes obvious.
Their subjects range from AI agents and enterprise software to healthcare, industrial automation, and brain-computer interfaces.
Follow their work for data-backed accounts of what is changing and questions such as whether AI companies will displace established platforms or become part of them.
Manlio Carrelli connects market signals to a strategic argument.
In one post, they trace Coinbase’s investments, partnerships, acquisitions, and hiring across five business areas to argue that the company is building an “everything exchange.”
In another, they challenge the idea that AI will simply replace SaaS, pointing to acquisitions and enterprise platform advantages before drawing on their contact-center experience to explain why labor costs may matter more than software costs.
They also turn market tracking into practical indicators, including startup success probabilities, exit probabilities, and commercial-maturity scores.
Their work pairs concrete company examples with broader interpretation, sometimes extending beyond deal analysis to careers, services, and human-machine interaction.
Manlio Carrelli writes about where enterprise AI money is moving.
Read more
Their public posts return to a handful of arguments: read hiring, investment and partnership signals before the news breaks; expect incumbents to buy AI startups rather than be replaced by them; measure AI opportunities against labor budgets rather than software budgets; treat partnerships as the structure that decides who owns a category; and expect work to reorganize around people who can orchestrate agents and exercise judgment. They build each case from named companies and specific numbers, and only occasionally step out of the analyst voice to state a personal view.
Predictive intelligence beats waiting for the news
Carrelli's central claim is that hiring, investment, partnership and funding signals reveal a company's direction months before announcements, giving an advantage to whoever reads them first.
Read more
They frame this as a measurable product claim, not a hunch: CB Insights' Mosaic success score is described as 4.7 times more predictive of startup success than backing from top-decile venture firms. Commercial Maturity scores are pitched to teams deciding which company is ready to partner with or buy.
Carrelli wrote the free AI Signal Tracker for personal use while trying to keep pace with AI, then opened it up.
It publishes success probabilities for individual companies, exit probabilities covering the next 24 months for acquisition and IPO candidates, commercial maturity scores, market maps and live private-market deal data.
Their Coinbase post applies the same method by hand, linking a $375M acquisition, broker-dealer hiring, Deribit's options infrastructure, a $7B Chainlink bridge and payment partnerships to a stated 2026 direction.
Incumbents are buying the disruptors, not being replaced by them
Carrelli argues the "AI kills SaaS" story has flipped: large software companies are absorbing AI startups instead of waiting to be disrupted, because they already own the customer relationships, workflows and data AI can build on.
Read more
They also flag where their own model is exposed: per-seat pricing looks fragile if agents replace human users. They expect incumbents to fight and acquire agent startups while running leaner with AI to protect cash flow.
Carrelli cites public SaaS acquisitions of AI startups projected to more than double year over year in 2025, 177 AI M&A deals in Q2 2025, and AI startups acquired around six years earlier than non-AI peers (seven years versus thirteen) at a median 17.1 times revenue. Workday's purchases for conversational recruiting, LLM development and contract intelligence illustrate how AI is being stitched into a platform rather than replacing it.
The budget AI actually replaces is labor, not software
A recurring priority in Carrelli's writing is sizing AI opportunities against staffing costs rather than software licences.
Read more
They argue that customer service winners will sell resolved tickets and higher conversion, not tools. The money at stake sits in labor budgets rather than software budgets, changing who can pay and how much.
Drawing on their time in contact center software, Carrelli observed that the real AI opportunity was eating into staffing costs 20 to 40 times the size of the software opportunity.
Customer success software was a market in the tens of billions, they note, while customer service labor runs into the hundreds of billions annually.
They apply the same logic to professional services, where the billable-hour model and junior-level work sit directly in the path of agents.
Partnerships, not products, decide who owns a new category
Carrelli treats partnership maps as the real structure of enterprise AI: no single company builds the whole stack, so hubs connecting models, data infrastructure, selling platforms and payment rails set standards and keep the customer relationship.
Read more
Each partnership they describe explains what both sides gain, not just that a deal exists.
Carrelli names OpenAI, Google, Shopify and Perplexity as the gravitational centers of agentic commerce, then shows incumbent payment networks staying central: Visa has partnered with Anthropic and IBM, while Mastercard and PayPal build comparable agentic payment systems. In industrial AI, they list Nvidia as the most-connected partner with seven relationships, alongside Microsoft, Google Cloud, Palantir, AWS and Qualcomm.
Work is being rebuilt around orchestration, and taste and judgment are the scarce assets
Carrelli argues that AI shifts people toward directing agent systems rather than doing the underlying tasks, and that the durable human advantages are judgment and taste — the bulwark against large volumes of AI-generated slop.
Read more
They also identify proprietary data and productized services as what separates firms that adapt from those that do not.
Carrelli cites McKinsey running 12,000 internal agents and Deloitte rolling out copilots to 100,000 employees.
They report a survey of more than 60 organizational leaders on where agents are already deployed: 41% in customer support, 41% in marketing, 39% in software development and 36% in sales.
A KPMG leader quoted in their post says new hires straight out of college are expected to perform at a senior associate level because AI is their teammate from day one.
Same teardown method, pointed at categories outside mainstream AI
Beyond market scoring, Carrelli applies the same structure to less-covered deep tech: describe the technical shift, name the companies involved, then state what the movement means for the sector.
Read more
The range shows that the method travels beyond mainstream AI.
In brain-computer interfaces, Carrelli tracks funding tripling to $867M in 2025, Neuralink's $650M raise at a $9.7B valuation, and three technical tracks running from precision surgery to AI-powered neural decoding. In industrial AI, they distinguish physical AI, industrial foundation models, edge computing and digital twins, and argue leaders must master orchestration or become commodity suppliers.
The brain-computer-interface post is where Carrelli most clearly steps out of the analyst register. They write that they understand the reluctance about blending human and machine but see the medical use cases as "unmitigated amazing news" — an explicit personal judgment rather than a reported finding.
A clear answer to what this work measures: early-stage companies ranked by a proprietary success score, each paired with an operating proof point.
Read more
It shows the method in its most concrete form.
Carrelli lists eight companies whose success-probability scores rose by 300 points or more, pairing each with a specific operating fact.
Firecrawl reached 350,000 developers on its web-scraping API and partnered with LangChain and Weaviate.
Leo AI reports 20,000 active engineers at Scania, HP, Siemens and Mobileye, with 96% accuracy on mechanical engineering tasks against generic tools.
Extend AI reached multi-million ARR and cash-flow positive status at Series A, with Square, Brex and Flatiron Health as customers.
Sola won Fortune 100 and AmLaw 100 deployments and raised $17.5M from a16z.
The case hinges on the closing comparison: the score is described as 4.7 times more predictive of startup success than backing from top-decile venture firms.
For brands planning a campaign
A ready-made format for a co-authored report on startup momentum: a short ranked list linking each proprietary score to an independently checkable operating metric.
Shows the core move in miniature: use scattered public signals to reconstruct a direction before it becomes obvious.
Read more
A useful way to assess whether the analytical claims hold together.
Starting from a single acquisition, Carrelli reconstructs Coinbase's roadmap across five fronts.
Tokenized securities: a $375M Echo acquisition plus broker-dealer hiring.
Prediction markets: Deribit's options infrastructure, early investments in The Clearing Company and Limitless Labs, the Kalshi partnership, and then an acquisition to own the stack.
Traditional finance rails: a crypto-as-a-service platform serving 240+ institutions, including JPMorgan and PNC, with a Standard Chartered partnership for prime services.
Cross-chain infrastructure: Base, investments in Axelar, Akash and LI.FI, and an exclusive Chainlink bridge for $7B in wrapped assets.
Payments: USDC at a $74B market cap alongside PayPal, Visa, Stripe and Shopify partnerships.
The pattern: invest early, validate through partnerships, acquire to own the stack.
For brands planning a campaign
Gives a report on enterprise M&A a structural spine — a repeatable three-step pattern — rather than a deal-by-deal recap.
A worked example of the central argument in Carrelli's public writing: incumbents are absorbing AI rather than being replaced by it, with a case built on numbers.
Read more
Carrelli counters the idea that AI agents will kill SaaS with acquisition data: public SaaS purchases of AI startups are projected to more than double year over year in 2025; AI M&A hit 177 deals in Q2 2025; and AI startups are being acquired roughly six years earlier than non-AI peers (seven years versus thirteen), at a median 17.1 times revenue.
Workday's acquisitions for conversational recruiting, LLM development and contract intelligence illustrate how AI is being stitched into a platform rather than replacing it.
They also flag the fragility of per-seat pricing if agents replace human users, drawing on firsthand experience in contact center software: the real AI opportunity was eating into staffing budgets 20 to 40 times the size of the software budget.
For brands planning a campaign
Supplies both an argumentative spine and a personal operator anecdote for a report on where enterprise AI money is moving.
A direct account of the product behind the analysis — what the scores, probabilities and maps are for, explained in plain language.
Read more
Carrelli describes building the tracker for personal use as a leader trying to keep pace, then releasing it free.
They outline its outputs: success probabilities for individual AI companies; exit probabilities over the next 24 months identifying acquisition targets and IPO candidates; commercial maturity scores for partnerships ready to scale; market maps; and live private-market deal and funding data.
Behind it, they say, are data collected directly from private companies and agents that automate analysis and send results into a CRM.
The tracker is designed to help users act before the news, on the premise that new models change things overnight and startups move from stealth to acquisition in months.
For brands planning a campaign
The clearest statement of what the firm sells; a partnership proposal can be built around the same scores rather than generic AI commentary.
Establishes firsthand operator experience, explaining why earlier generations of automation failed and why this one is different — the credibility behind later market claims.
Read more
Carrelli recounts deploying natural-language systems in 2018 that classified intent and followed decision trees, but broke as soon as a customer said something slightly unexpected.
The goal, they write, was never better bots, but fewer humans providing better service instantly.
They contrast that with current revenue growth at companies such as Sierra, Crescendo, Decagon and Chatbase, and size the stakes: customer success software was a market in the tens of billions, while customer service labor runs into the hundreds of billions annually.
Their conclusion: winners sell resolved tickets and higher conversion, not tools; model makers will go direct; and control of the customer experience is the defensible layer.
For brands planning a campaign
Shows how Carrelli can open a piece with an operator's story and land on a market-structure claim, a fit for a co-authored report targeting enterprise buyers.
Takes the same thesis from software markets to the way organizations and jobs are changing.
Read more
Read it to see whether the argument holds beyond dealmaking.
Carrelli quotes a KPMG leader saying new graduates are expected to perform at senior associate level because AI is a teammate from day one, and cites McKinsey running 12,000 internal agents and Deloitte rolling out copilots to 100,000 employees.
They report a survey of more than 60 organizational leaders on where agents are already deployed: 41% in customer support, 41% in marketing, 39% in software development and 36% in sales.
Five adjustments follow: orchestration rather than execution, proprietary data as differentiation, products over projects, human-AI teams, and taste and judgment as the scarcest assets.
For brands planning a campaign
Broadens a partnership beyond buyers to enterprise leaders and gives a co-authored report a workforce section beyond tooling.
A compact case study in how partnerships, not products, decide which companies own a new category.
Read more
Useful for judging the partnership-overview format.
Carrelli names four companies as the gravitational centers of AI commerce.
OpenAI has 16 partnerships since 2024 plus a $1.1B Statsig acquisition.
Google is embedding assistants through brand partners including Home Depot, Behr and Alaska Airlines.
Shopify is positioned at checkout through Shop Pay integrations with OpenAI and Perplexity.
Perplexity is the emerging integrator across discovery and search.
They identify the four capabilities no single company covers — models, data infrastructure, selling platforms and payment rails — and note that Visa has partnered with Anthropic and IBM while Mastercard and PayPal build similar agentic payment systems, keeping incumbent payment networks central.
For brands planning a campaign
A template for a partnership-ecosystem brief: name the hubs, identify the missing layers and explain why no one can build it all alone.
The most direct example of how a commercial partnership gets announced and framed — what the deal is, who it serves, and what claim it is meant to support.
Read more
Carrelli announces a partnership that puts commercial maturity scores inside Clay's go-to-market workflows, crediting Varun Anand and the Clay team for moving quickly.
They name the audience already using the data: business development and partnership teams at large technology companies, banks and professional services firms, plus M&A teams looking for targets early.
The claim they attach to it is that teams using predictive intelligence close more deals, faster, than teams working without it, framing the shift as a move from reacting to what already happened to anticipating what comes next.
For brands planning a campaign
Shows how Carrelli turns a distribution deal into a narrative about a market shift rather than a product announcement — relevant when a partnership includes joint marketing.
Explains why this category is so hard to evaluate from the buyer's side, which is exactly the problem a market-intelligence partnership exists to solve.
Read more
Carrelli counts the competitive density in vertical categories: more than 100 coding agents, over 100 customer service agents, over 100 legal agents, and dozens more in recruiting, sales, finance, IT and supply chain.
Their argument is that when an agent executes an outcome across systems of record, the software underneath starts to look interchangeable, pressuring SaaS revenue and gross margins — uncertainty already visible in how public markets price software businesses.
They expect incumbents to both fight and acquire agent startups while running leaner with AI to protect cash flow, and to keep watching private markets for the next experiments.
For brands planning a campaign
A supply-side framing for a report section on why enterprise buyers struggle to compare vendors in a crowded category.
Strategy maps built from hiring, investment, and partnership signals3
AI's effect on work, roles, and professional services3
Category teardowns of emerging tech2
Momentum scores for private AI companies
Carrelli builds recurring content around CB Insights' Mosaic success-probability score and Commercial Maturity score, naming individual companies and the numbers behind them.
Read more
One post lists eight companies that each jumped 300-plus Mosaic points and describes their developer counts, enterprise customers, and funding rounds; another frames the same scores as a way to see which AI companies are winning before the news does. The scoring method is explicit, and every company entry includes a concrete commercial fact.
Company-by-company breakdowns anchored to a named score, specific round sizes, and customer names.
For brands planning a campaign
A co-authored report on AI startup momentum can lead with the scoring method, then use named companies and their numbers as the body. Carrelli has repeatedly used this format, building authority through an explainable data pipeline rather than opinion.
Carrelli argues that the 'AI kills SaaS' narrative has flipped: large software incumbents are acquiring AI startups instead of waiting to be disrupted.
Read more
Their posts include M&A counts, exit timing (AI startups acquired roughly six years earlier than non-AI peers), median revenue multiples, and Salesforce's acquisition history. They ground the trend in their own contact-center experience, where the budget being replaced was for labor, not software.
For brands planning a campaign
For an enterprise M&A trends report, Carrelli offers both aggregate deal statistics and the operating logic behind them—why incumbents buy and which budget lines move—so a section can go beyond deal counts to consequences.
Strategy maps built from hiring, investment, and partnership signals
Carrelli assembles forward-looking 'strategy maps' that connect acquisitions, investments, partnerships, and hiring data to predict where a company is headed years out.
Read more
The Coinbase post links five investment and acquisition threads to stated 2026 directions; the agentic-commerce post names four companies as the gravitational centers and describes payment networks quietly preparing behind the scenes. They frame the method as seeing direction before it becomes obvious.
Layered evidence (investments plus hiring plus partnerships) presented as numbered 'what it signals' threads.
For brands planning a campaign
This approach makes non-obvious inferences auditable by layering separate signal types into a stated direction, rather than presenting a prediction as a guess.
Carrelli maps partnerships as the organizing structure of enterprise AI.
Read more
They list Nvidia, Microsoft, Google Cloud, Palantir, AWS, and Qualcomm as partners industrial leaders scale with; explain why no single company builds the full agentic-commerce stack; and describe a data partnership that places predictive signals inside GTM workflows. Each case explains what the partnering companies gain, not just that the partnership exists.
For brands planning a campaign
For a brand evaluating a thought-leadership partner on AI momentum and M&A, this shows Carrelli can write partner-ecosystem sections that explain strategic rationale and switching costs, rather than deliver trade-press summaries.
AI's effect on work, roles, and professional services
Carrelli returns to labor economics: the billable-hour model under pressure, AI agents taking over junior-level work, and deployment rates across customer support, marketing, software development, and sales.
Read more
They use a named practitioner quote, company-level agent counts (McKinsey's 12,000 internal agents; Deloitte copilots for 100,000 employees), and a five-part framework for how firms and individuals adapt. The framework includes their view that taste and judgment are the defense against AI-generated slop.
The contact-center post is a personal recollection from 2018, not a data post, and reaches the same labor-replacement conclusion through Carrelli's own operating history.
For brands planning a campaign
A co-authored report can add a human-consequence section with quantified adoption and named firm examples. The material also shows Carrelli will state a judgment about what stays valuable, rather than only report data.
Carrelli applies the same structure to categories outside mainstream AI coverage: brain-computer interfaces (funding tripled to $867M, Neuralink at a $9.7B valuation, three technical tracks from surgery to neural decoding) and industrial AI (physical AI, industrial foundation models, edge computing, digital twins).
Read more
Each post explains the technical shift, names the companies, and closes with what the movement means for the sector.
The brain-computer-interface post includes Carrelli's own stance—they call the medical use cases 'unmitigated amazing news'—and is the clearest example in these posts of editorializing rather than reporting.
For brands planning a campaign
The same repeatable teardown format extends beyond core AI-market list posts to less-covered deep-tech categories.
Carrelli writes from inside CB Insights, the market-intelligence firm behind the Mosaic success-probability scores, exit probabilities, and AI Signal Tracker they describe as “our data” and “we built.”
Read more
They directly announce one brand relationship: a CB Insights partnership with Clay, which pipes CB Insights signals into Clay’s go-to-market workflows.
The other companies appearing in their posts — Firecrawl, Leo AI, Siemens, Coinbase, OpenAI, Shopify — are subjects of analysis, not sponsors or clients.
They also mention working in contact center software in 2018, without naming the employer.
CB Insights
Employment
Carrelli posts as a CB Insights member, not an outside commentator: they call its Mosaic success-probability score and AI Signal Tracker “our data,” say “we built this AI Signal Tracker,” and credit CB Insights as the data source behind market analysis.
Carrelli announced a CB Insights partnership with Clay, making CB Insights signals such as the Commercial Maturity score available in Clay’s go-to-market workflows.
Read more
They thanked Clay’s co-founder and team for moving quickly on the partnership.
Manlio Carrelli's posts form a running argument about what money, hiring and partnership deals reveal about enterprise AI before announcements land.
Read more
Start with the two momentum pieces — the list of eight private AI companies that each gained more than 300 points on CB Insights' Mosaic success score, and the free AI Signal Tracker post — which introduce the scores and sources they reuse elsewhere.
Then read the strategy maps, especially the Coinbase post, which links a $375M acquisition, broker-dealer hiring, options infrastructure and a $7B bridge to a stated 2026 direction.
It shows how Carrelli builds a forward view from separate signal types rather than a single announcement.
The incumbent-software posts argue that large software companies are buying AI startups instead of being replaced by them, citing deal counts, exit timing and revenue multiples.
Two recurring conclusions follow: AI replaces labor budgets, not software licences, and partnerships rather than products decide who owns a new category.
The brain-computer-interface and industrial AI posts apply the same method to less-covered categories.
Marketers seeking market commentary tied to named companies, funding figures and a stated scoring method will find a consistent body of work; those seeking hands-on tutorials, technical walkthroughs or personal commentary will find less.
The posts carry no dates, so ask for publication dates if chronology matters.
Both write for business and technology decision-makers about which enterprise technologies are gaining ground and why.
Read more
Atkinson's award roundup names companies across IoT, edge AI, data intelligence and emerging tech, while their industrial-controls post argues that connected devices should never have been left exposed. Carrelli applies the same instinct to explain which companies are gaining momentum and what that means for buyers.
Atkinson's posts centre on industry recognition, standards bodies such as CISA and NIST, and security hygiene for operational technology, drawing on their work as a marketer and awards organiser. Carrelli builds cases from private-market financial signals — success scores, funding rounds and acquisition multiples — giving readers who follow both ecosystem coverage from Atkinson and investment-led inference from Carrelli.
Both translate enterprise AI and robotics announcements into business implications.
Read more
Bourouies's Hyundai post explains a $3B robotics investment, the edge AI chip built with DEEPX and the digital-twin validation step, then states what the plan implies for the industrial landscape. That structure mirrors Carrelli's practice of ending each breakdown with the strategic consequence, not just the news.
Bourouies focuses on one company or decision at a time — Hyundai's robotics programme, Anthropic's restricted release of a vulnerability-finding model — and leans on narrative explanation. Carrelli layers several signal types in a single post and attaches proprietary scores, deal counts and revenue multiples, building an argument from accumulated signals rather than one story.
Both explain what current AI systems can and cannot do, and where the next layer of capability may come from.
Read more
Daali's world-models essay covers persistent memory, planning and physical grounding as limits of language models, while the agents piece argues that agents should be treated as digital employees needing goals and feedback. Carrelli covers similar territory, writing that work is reorganising around people who orchestrate agents.
Daali writes long, research-based teaching explainers on topics such as LeCun's JEPA, developmental psychology and sensorimotor learning, and invites readers into a community discussion. Carrelli focuses on commercial commentary: named companies, funding figures and scoring products for readers making investment, partnership or buying decisions.
The LinkedIn profile listed 9,251 followers on 8 October 2026.
Read more
No engagement rates, impressions, follower growth, audience demographics or benchmarks are published, so the figure shows the size of the following, not who is in it or how they respond.
The posts are short market-analysis pieces: company-by-company reads on AI momentum, M&A and partnership activity, and product news.
They’re aimed at readers who follow enterprise technology and dealmaking rather than a general audience.
Recurring subjects—including AI startup tracking, incumbent-versus-startup competition, agent infrastructure and enterprise buying—suggest overlap with strategy, corporate development and market-intelligence roles.
The topic is a close fit for the proposed co-authored report on AI startup momentum and enterprise M&A trends, but the published work doesn’t establish whether readers are the decision-makers the report needs to reach.
Ask for LinkedIn analytics from the last 12 months, including impressions and engagement on AI-momentum and M&A posts, follower growth, and follower breakdowns by industry and seniority.
If available, request an audience survey covering job function, company type and seniority.
LinkedIn followers
9,251
Follower count recorded on the LinkedIn account that published the reviewed posts (linkedin.com/in/mcarrelli) · 2026-10-08
No sponsorship policy, media kit, rate card, or partnership inquiry route appears in Manlio Carrelli’s posts, and they do not describe an individual sponsorship format.
Read more
Instead, the posts follow a consistent company-publication pattern: posts on AI company momentum, enterprise software economics, and market maps end with a link to a free CB Insights tracker or a full report.
One post announces a company-level partnership between CB Insights and Clay, embedding CB Insights predictive scores into Clay’s sales workflows.
That is a corporate data and go-to-market partnership, not a creator sponsorship.
For a campaign aimed at readers interested in AI startup momentum and enterprise M&A trends, the posts point to a company-to-company content arrangement with CB Insights rather than a personal sponsorship deal; the closest precedent is the existing CB Insights report-and-tracker publishing model.
The posts do not specify individual deliverables, an approval process, exclusivity terms, or disclosure practices.
Ask CB Insights directly about budget, turnaround, and editorial control rather than assuming personal terms.
Strong fit. This is the closest match to work they already do repeatedly and publicly.
Read more
Brief
A B2B AI market intelligence firm wants to assess Manlio Carrelli for a thought leadership partnership and proposes a co-authored report on AI startup momentum and enterprise M&A trends.
Reasons
The momentum half of the brief is already a working format. One post ranks eight companies whose success-probability score rose 300 points or more, and backs each entry with an independently checkable operating fact rather than an adjective: Firecrawl has 350,000 developers on its scraping API and partnerships with LangChain and Weaviate; Leo AI has 20,000 active engineers across Scania, HP, Siemens and Mobileye; Extend AI has multi-million ARR, is cash-flow positive at Series A, and counts Square, Brex and Flatiron Health as customers; Sola has Fortune 100 and AmLaw 100 deployments and a $17.5M a16z round. The closing comparison gives the method its authority: the score is presented as 4.7 times more predictive of startup success than backing from top-decile venture firms.
The M&A half is equally developed. Carrelli reports that public software incumbents' purchases of AI startups are projected to more than double year over year, that AI M&A hit 177 deals in Q2 2025, that AI startups are acquired roughly six years earlier than non-AI peers, and that median revenue multiples sit near 17.1 times. They name Salesforce's acquisition history and Workday's buys in conversational recruiting, LLM development and contract intelligence, then argue the direction of travel rather than simply listing deals.
The two halves can be joined by a pattern they have already articulated: invest early, validate through partnerships, then acquire to own the stack. The Coinbase piece runs that pattern across five fronts, from a $375M acquisition and broker-dealer hiring to prediction markets, prime services, cross-chain infrastructure and a $74B stablecoin position. Applying the same reconstruction to a set of AI categories is a natural extension, not a new skill.
The proposed deliverable has a precedent in their publishing. Posts end with a link to a free tracker or a full report, making a co-authored report a familiar format for this author and audience. One post documents embedding a commercial-maturity score inside a partner's sales workflows, showing how a data partnership can reach company buyers.
Subject-matter audience alignment is plausible: the recurring subjects are enterprise technology, dealmaking and market intelligence, all relevant to a firm selling AI market intelligence.
Open questions
Ask for audience analytics to assess reach and engagement.
Ask directly about budget, turnaround, editorial control, approval process, exclusivity and disclosure practices; the posts do not establish these terms.
When to choose a specialist
Poor fit. The mismatch is format, posture and measurement, not topic.
Read more
Brief
A developer-tools vendor wants a hands-on product tutorial series: recorded walkthroughs and live demos, with a guaranteed reach commitment and demographic reporting measured against paid-media targets.
Reasons
The posts show no video, recording or tutorial format; they are short written analyses built around market data. Ask for samples of recorded walkthroughs or live demos before committing to these deliverables.
The register is market and strategy commentary, not practitioner instruction. The posts explain why a category is moving, who is winning and what deal activity signals. They do not teach readers how to use a tool, work through a setup or compare features — the practical instruction a developer-tools buyer expects from a tutorial.
The commercial pattern is company publishing, not independent creator review. Posts on AI momentum, market maps and enterprise software link to a free tracker or full report, and one announces a distribution deal that puts commercial-maturity scores inside a partner's workflows. A vendor seeking hands-on, comparative tool coverage would need to agree on an editorial posture not demonstrated here; the closest example instead promotes the product family underpinning the analysis.
Ask for impressions, engagement rates, follower growth and audience breakdowns by industry, function and seniority before setting reach or demographic targets.
Ask for a rate card, media kit, exclusivity terms and branded-content disclosure practices before negotiating a deal based on reach guarantees.
What would change this
The posts show no recorded or instructional formats, hands-on demonstrations of third-party products or audience analytics. Ask for examples of these before considering a campaign built around them.
A worked campaign example
Assets
A co-authored long-form report paired with a lightweight free tracker or scored index would match the publishing model already in use: a short post carrying the argument, a link to the full report, and a free data entry point that draws readers into the firm's pipeline.
Brief
Co-authored report on AI startup momentum and enterprise M&A trends for a B2B AI market intelligence firm.
Campaign angle
Position the report as a decision tool rather than a trend summary. The selling line is anticipation: the analysis aims to identify acquisition targets and partnership candidates before announcements. The recurring phrase in the reviewed work — act before the news — is a natural campaign hook.
Central claim
Large software incumbents are absorbing AI capability rather than being replaced by it. Public signals make the pattern of absorption — invest early, validate through partnerships, acquire to own the stack — legible months before announcements.
Structure
Content
Open on the acquisition math: purchases of AI startups by public software incumbents projected to more than double year over year, 177 deals in Q2 2025, AI startups acquired roughly six years earlier than non-AI peers, and median multiples near 17.1 times revenue. State the conclusion in the first paragraph: incumbents are buying time, talent and technology rather than waiting to be disrupted.
Section
The claim, up front
Content
Lay out the three-step sequence that recurs in the posts: early minority investment, validation through a partnership or customer relationship, then acquisition to own the layer. Trace five threads through acquisitions, investments, partnerships and hiring at one company, so readers can see the inference and judge it.
Section
The pattern, named and shown
Content
Rank a short list of candidate companies using the success-probability and commercial-maturity framing, and give every entry one independently checkable operating fact: developer or engineer counts, named enterprise customers, revenue status at the last round, or partnership depth. Avoid follower-count or funding-total rankings; the point is which companies are ready to partner or be bought.
Section
The Shortlist and the Score
Content
Add a section on budget pools. The addressable spend often comes from the labor line rather than the software line: contact-center staffing budgets ran 20 to 40 times the size of software budgets. This is the section most likely to change a reader’s mind about category sizing.
Section
Where the Money Comes From
Content
State plainly what would break the thesis: per-seat pricing comes under pressure if agents replace a meaningful share of human users, incumbent self-disruption stalls, or private-market supply outpaces what enterprises can evaluate. Naming these failure conditions makes the forecast testable rather than promotional.
Section
The counter-case
Content
Close with a short list of names to watch, the signals that would move each one, and a time horizon for the next update.
Section
What to do next
What to settle first
Ask for the deliverable list and length, who holds editorial control, the approval and disclosure process, exclusivity terms, the timeline, and audience analytics to judge reach.
Working title
What Incumbents Buy Next: Where AI Deal Activity Is Heading Before It Is Obvious
When a report should focus on industrial and infrastructure sectors—IoT, edge AI, connectivity, critical-infrastructure security—and offer category-level vendor recognition rather than a ranked momentum score.
Read more
The two posts show that pairing directly: one announces an awards program with categories spanning connected building, fleet and asset management, AI sensors, semiconductors, agentic AI and brain-computer interfaces, and names dozens of vendors across IoT, business tech and emerging tech.
The other urges operators of water, waste and utility systems to secure programmable logic controllers, pointing readers to CISA's cross-sector performance goals and NIST's control-system security guidance.
For a report aimed at utilities, telcos or industrial operators, that sector taxonomy and compliance guidance is closer to readers' daily concerns than deal-multiple analysis.
Neither post draws on deal or funding data.
One lists vendor award recipients; the other advocates for control-system security.
Neither ranks companies by likelihood of success or readiness for acquisition or partnership, or supplies the transaction figures—deal counts, revenue multiples, acquisition timing—that an enterprise M&A report needs.
A firm seeking measured momentum signals would need to provide that dataset and accept a sector- and compliance-first framing.
When a report needs readable explainers of individual company moves in physical AI and AI security, and readers want to know what a company has just committed to and what it implies—not which private companies come next.
Read more
One post examines Hyundai's robotics and physical-AI plan: Boston Dynamics Atlas units slated for a Georgia plant, an edge AI chip co-developed with DEEPX, a $3B investment alongside NVIDIA and 50,000 Blackwell GPUs, and digital-twin validation before real-world rollout.
It draws out the industrial and workforce implications for a business audience.
The other explains Anthropic's restricted release of a vulnerability-detection model and the defensive-security consortium built around it, setting out the dual-use tradeoff in plain terms.
The writing explains announced plans rather than analyzing market data.
The two posts neither rank companies nor score momentum or aggregate deals and funding.
A co-authored momentum report would need the partner to provide the dataset and analytical method, while this voice narrates individual announcements.
The sample is narrow: two company cases, one in robotics and one in AI security.
When the deliverable should teach the concepts behind the market rather than cover the market itself—for example, a companion explainer for readers who need to understand architectures and capability limits before judging vendor claims.
Read more
The AI Minds posts reviewed do just that: one explains why current language models struggle with persistent memory, generalization to new environments and physical grounding, then introduces world models and JEPA as an alternative direction, anchored by LeCun's framing. The other offers a working framework for AI agents—tool use, planning, reflection and memory—and argues for treating them as digital employees that need goals, frameworks and feedback.
This educational, conceptual writing explains mechanisms, offers a point of view and invites readers into a community rather than directing them to a data product.
The posts include no company-by-company rankings, deal or funding analysis, or enterprise M&A material.
For a report on startup momentum and acquisition trends, this voice can strengthen the explainer layer but cannot replace the data-led core.
Manlio Carrelli's 13 LinkedIn posts cover AI company momentum scores, incumbent software M&A, agentic commerce partnerships, professional services and the future of work, brain-computer interfaces, a payments company's acquisition strategy and industrial AI.
They are full post texts rather than titles or excerpts.
The posts have no visible publication dates, so the page is organised by theme rather than as a chronology.
The posts do not mention any book, show or signature project; a full career history beyond one reference to contact center software work in 2018; video or audio material; audience analytics; or sponsorship or paid-client terms.
Three other creators are referenced for comparison.