Jeff Epstein is an investor at BVP and a former CFO who writes about how finance teams can use AI, measure its business impact, and set controls for AI agents.
Jeff Epstein helps finance leaders test what AI changed in finance work and what controls it needs.
Best for
B2B finance software campaign for CFOs evaluating AI in finance operations.
Not for
Consumer or small-business accounting apps seeking product demos, guaranteed attendance, or paid media placement.
Key takeaways
A faster close does not necessarily mean a faster business.
Define what AI agents may do before naming their permissions.
The real cost of AI in finance is time people spend checking it.
Check first: Ask what paid partnership terms are available and whether a paid partnership route exists.
Main ideas
"How are you using AI" has been replaced by "what actually changed" · Define what agents may do before naming their permissions · A faster close is not a faster operating model
Jeff Epstein is an investor at BVP and a former CFO at Oracle who writes about putting AI to work in finance.
Their posts ask what changes beyond the technology itself: business results, decision-making, and accountability.
Their perspective is useful to finance leaders weighing AI’s practical value and the controls it needs.
Epstein makes broad AI claims concrete by tracing their consequences through finance work.
In one post, they argue that a faster close does not necessarily mean a faster business: if pricing and board decisions still follow the old cadence, automation has sped up reporting without changing the operating model.
In another, they turn agentic AI risk into a governance test: who owns the decision, what may an agent do, and when must it stop and escalate to a person?
They also distinguish measurable outcomes from forecasts, questioning whether anyone can reliably predict a sweeping estimate of AI’s potential profits.
For readers evaluating AI in finance, that combination of operational detail, business measures, and attention to limits makes their work worth following.
Jeff Epstein writes about AI in finance as someone who bought technology as a CFO and now questions founders about it as an investor.
Read more
Across five LinkedIn posts, the recurring argument is that the finance conversation about AI has moved past adoption and into measurement and control: what changed in the business, who is accountable when an agent acts, and whether faster reporting actually changed how decisions get made. Their posts use survey numbers, ask practical questions instead of offering answers, and repeatedly point finance leaders toward comparing notes in person rather than learning from vendor demos.
"How are you using AI" has been replaced by "what actually changed"
Epstein argues that a couple of years ago, the finance question was mostly how a team was using AI; now it is what actually changed as a result.
Read more
The test they set out is deliberately financial: did the team get more productive, did the economics change, are decisions faster or better, and can any of it be measured?
The processes they name are forecasting, reporting, close, collections and analysis; in a later post, the same test extends to margins, pricing and retention.
They describe this from both sides of the table: proving technology investments were worth what was paid for them as a CFO, and now asking founders to do the same as an investor.
The question list itself is the example: Epstein asks, "If you're a finance team using AI in forecasting, reporting, close, collections, or analysis, what happened?"
They follow with questions about whether the team got more productive, whether the economics changed, whether decisions are faster or better, and whether any of it can be measured.
In a separate post, they apply the same standard to margins, pricing, retention and collections, and ask whether AI lets a company do something it could not do profitably before.
In one post, Epstein states plainly that they changed their mind.
They once saw the CFO's job with AI as making sure the company was not wasting money on it; now they argue the question is what AI is doing to the business — margins, pricing, retention and collections.
Their earlier focus on proving technology spend was worth it as a CFO carries into how they question founders as an investor.
Define what agents may do before naming their permissions
Epstein's governance argument starts with a gap: finance controls were designed around people — someone approves a payment, someone owns the close, someone signs off on the numbers — but no equivalent exists for software agents that touch cash, close, collections, reporting, procurement or pricing and make decisions.
Read more
Their position is not to slow or stop AI, but to set boundaries first: who owns the decision, what agents may act on, and what must be stopped and escalated to a person. In a related post, they make the same point: finance has to help decide where machines can act on their own and where human judgment still matters.
Epstein opens with a survey finding that 23% of finance leaders said that if a significant AI agent error happened overnight, it would not be clear by 9 AM who was accountable. They contrast that with the human controls finance already knows — who approves a payment, who owns the close, who signs off on the numbers — and land on the rule, "Think human before you name the agent's permissions."
Epstein stacks automation gains against the state of finance's underlying data and controls to argue that finance is getting better at producing information without changing how the business decides.
Read more
If the numbers are ready Tuesday instead of Friday but pricing decisions still wait for the monthly meeting, or the close is automated while the board cadence stays the same, the old operating model has simply been sped up. The question they say they are most interested in is what happens when finance stops being a function that reports and starts operating proactively.
Consero's 2026 CFO survey puts the average monthly close at 6.1 days, down from 8.9 days in 2022, and says 87% of CFOs expect fully automated journal entries within 12 months.
Against that, Epstein cites figures showing that only 4% of CFOs say they have a single source of truth for their data and 96% are still verifying AI outputs every day.
They conclude: "We've just made the old one faster."
The real cost of AI in finance is the time people spend checking it
Epstein's most concrete operational point is about rework.
Read more
They name where AI is showing up — billing, reconciliation, forecasting, reporting, close, collections and controls — then describe the awkward position finance teams are in: asked to help lead the company's AI conversation while still working out where AI works inside finance.
The questions they raise are practical: what people are actually using, what is working, what is not, where a person is still needed, and where one can finally be taken out of the process.
They also say directly that there is a limit to what can be learned from vendor demos and LinkedIn posts.
A survey finding that finance teams spend 26% of their week fact-checking or correcting AI output — more than a full day — is the anchor. Epstein pairs it with a plain admission: "There's only so much you can learn from vendor demos and LinkedIn posts."
Peer conversation is how finance leaders should evaluate AI
Every post is built around the same in-person event, Finance Exchange in San Francisco on October 15.
Read more
Epstein says finance leaders need to sit down with each other and compare what is actually happening, which demo-heavy formats cannot deliver.
The posts promote the event with an application or registration link and report demand — 200+ finance leaders registered in one, oversubscribed in another.
The tone is informal alongside the numbers, including the sign-off, "let's go to the moon with finance."
One post names Priya Rajan as a co-speaker for a session on what finance leaders are doing with AI today, where it creates leverage and where it does not; another says the event is oversubscribed and offers to fit people in.
Zenskar appears in the event branding and activity links.
Ask for sponsorship terms and fees before booking.
This is the clearest statement of Epstein’s argument and why they make it: AI has to be defended in financial terms, not demonstrated.
Read more
A newcomer who reads only one post should read this one.
Epstein frames this post around having sat on both sides of technology spending: buying it as a CFO at Oracle and other companies, and now asking founders to justify it as an investor at BVP.
The argument is that the finance question about AI has moved on from “how are you using it?” to “what actually changed?”
The post lists what a finance team should be able to answer: did the team get more productive, did the economics change, are decisions faster or better, and can any of it be measured?
It ends by promoting a Finance Exchange session in San Francisco on October 15 with Priya Rajan, covering what finance leaders are doing with AI today, where it creates leverage and where it does not.
For brands planning a campaign
A finance-software vendor targeting CFOs can borrow the framing directly: the session pitch becomes “bring your measured results, not a vendor demo,” which fits an executive roundtable better than a product webinar.
It shows how Epstein’s position has developed and how they handle statistics — using them as direction rather than proof.
Read more
Useful for readers judging whether the argument is grounded.
Epstein writes that when AI first became a serious finance conversation, the CFO’s job seemed to be stopping waste on it.
That view has changed.
The replacement question is what AI is doing to the business — margins, pricing, retention, collections, or whether it lets the company do something it could not profitably do before.
The post cites two third-party figures: 74% of business leaders expect to deploy agentic AI within two years, and Bain estimates $4.7 trillion in global profit is at stake between 2025 and 2035.
Epstein says outright that nobody can really predict a number like that, while accepting the underlying direction.
For brands planning a campaign
Vendors should avoid putting an exact ROI projection in Epstein’s mouth; the statistic that travels with the argument is the split between where machines can act alone and where human judgment stays.
This is the most concrete governance argument in the material and turns a survey figure into a decision rule a finance team could apply the same week.
Read more
The post opens with a single number from a CFO Monday Brief survey: 23% of finance leaders said that after a significant AI agent error overnight, it would not be clear by 9 AM who was accountable.
Epstein contrasts that with controls built around people — who approves a payment, who owns the close, who signs off on the numbers — and asks what happens when agents touch cash, close, collections, reporting, procurement, or pricing and make a decision.
The stated position is not to slow AI down but to define boundaries first: who owns the decision, what agents may do, what must be escalated to a human.
It closes with the line "Think human before you name the agent's permissions."
For brands planning a campaign
A controls, audit, or workflow vendor can build a roundtable segment around the escalation question — what agents may do unattended versus what stops for a person — rather than around AI capability.
It is the best example of how Epstein uses numbers to argue rather than decorate, and answers the obvious objection to AI in finance: speed alone is not the point.
Read more
Epstein sets Consero’s 2026 CFO survey numbers against each other: the average monthly close is 6.1 days, down from 8.9 in 2022, and 87% of CFOs expect fully automated journal entries within 12 months.
On the other side, only 4% say they have a single source of truth for their data, and 96% are still verifying AI outputs every day.
The distinction is easy to miss: producing financial information faster is not the same as deciding faster.
If pricing decisions still wait for a monthly meeting, or the board cadence is unchanged, the operating model has not changed — it has only accelerated.
The post also notes that more than 200 finance leaders registered for the October 15 Finance Exchange.
For brands planning a campaign
The “faster close versus faster decisions” contrast is a ready-made webinar hook for close, consolidation, or planning software, inviting CFOs to describe where their own decision cadence is still stuck.
It names the specific functional areas a brand would care about and quantifies the rework AI creates, a practical detail most commentary omits.
Read more
Epstein lists where AI is showing up in finance — billing, reconciliation, forecasting, reporting, close, collections, controls — then cites a survey finding that finance teams spend 26% of the week fact-checking or correcting AI output, more than a full day.
The tension is that finance is being asked to lead the company’s AI conversation while still working out where AI works inside finance itself.
The post closes with questions rather than answers: what are people actually using, what is working, what is not, where is a person still needed, and where can one finally be taken out of the process?
It also notes the event is oversubscribed and takes a mild jab at vendor demos and LinkedIn posts as sources of learning.
For brands planning a campaign
An honest framing for a vendor: acknowledge the 26% correction time as the buyer’s real objection, then position the roundtable as a place to compare what reduces it. Epstein is skeptical of demos.
From "how are you using AI" to what actually changed2
Agent accountability: who owns the decision by 9 AM2
A faster close is not a faster operating model1
The verification load is still a person's job1
From "how are you using AI" to what actually changed
Epstein's most consistent argument is that AI in finance has moved past the usage question.
Read more
In one post, they write that a couple of years ago the question was mostly "how are you using AI?" and now it is "OK, what actually changed?" — did the team get more productive, did the economics change, are decisions faster or better, and can any of it be measured.
The same post names the finance processes in scope: forecasting, reporting, close, collections, analysis.
A later post applies the same test to margins, pricing, retention and collections, and asks whether AI lets a company do something it could not do profitably before.
Epstein frames this from both sides of the table, describing a CFO background proving technology investments were worth what was paid for them and a current role as an investor asking founders the same question.
Reframes an old question into a sharper one, then lists concrete processes and outcome measures so the abstraction stays checkable.
For brands planning a campaign
A vendor pitching AI in finance operations can borrow this framing directly: lead with measurable before-and-after outcomes per process rather than product capability. For the proposed roundtable, a session built around "what changed and how do you know" gives Epstein a question they already ask publicly and gives the audience a way to compare notes.
Agent accountability: who owns the decision by 9 AM
Epstein opens one post with a single number: 23% of finance leaders surveyed said that if a significant AI agent error happened overnight, it would not be clear who was accountable by 9 AM.
Read more
They argue that governance structures were built around humans — someone approves a payment, someone owns the close, someone signs off on the numbers — and that nothing equivalent exists for agents touching cash, close, collections, reporting, procurement or pricing.
Their conclusion is not to slow down or stop AI but to define boundaries before agents act: who owns the decision, what agents may do, and what must be stopped and escalated to a person.
The line they land on is "think human before you name the agent's permissions," and a separate post makes the same point: finance has to help decide where machines can act on their own and where human judgment still matters.
Leads with one survey statistic, builds the gap with a list of human-owned controls, then closes on a memorable rule for permissions.
For brands planning a campaign
This is the sharpest available hook for a CFO audience evaluating agentic AI. A roundtable or webinar anchored on accountability and escalation rules gives attendees a decision framework rather than a product tour, and lets a sponsor position controls and auditability as the topic, not the sales pitch.
Consero's 2026 CFO survey puts the average monthly close at 6.1 days, down from 8.9 days in 2022, and 87% of CFOs expect fully automated journal entries within 12 months.
Against that: only 4% of CFOs say they have a single source of truth for their data, and 96% are still verifying AI outputs every day.
Their argument is that finance is getting good at producing information faster without changing how the business decides — if the numbers are ready Tuesday instead of Friday but pricing decisions still wait for the monthly meeting, or the close is automated while the board cadence stays the same, the old operating model has simply been sped up.
The question they say they are most interested in is what happens when finance stops reporting and starts operating proactively.
Contrast structure that sets a strong automation statistic against a lagging control statistic, then reframes both as an operating-model question.
For brands planning a campaign
Useful for a vendor that wants to sell decision speed rather than cycle-time reduction. It gives a CFO audience a reason to examine what happens downstream of automation and suggests a session exercise around which decisions would actually move if the reporting cadence did.
Epstein's most concrete operational observation is about cleanup.
Read more
One post cites a survey finding that finance teams spend 26% of their week fact-checking or correcting AI output — more than a full day — and describes the position finance is in: asked to help lead the company's AI conversation while still working out where AI actually works inside finance itself.
The questions they set out are practical rather than strategic: what are people actually using, what is working, what is not, where do you still need a person, and where can you finally take a person out of the process.
They also say there is a limit to what can be learned from vendor demos and LinkedIn posts.
Names a specific, quantified friction point, states the audience's awkward position plainly, then converts it into a short list of practical questions.
For brands planning a campaign
A vendor that acknowledges the verification burden and can show where a person genuinely comes out of the loop will read as more credible to this audience than one leading with autonomy claims. This makes a natural peer-roundtable prompt: where are humans still required?
All five posts center on the same in-person event: Finance Exchange in San Francisco on October 15.
Read more
They promote it with an apply-or-register link, report participation (200+ finance leaders registered in one; oversubscribed in another), and reference co-speakers.
Zenskar's name appears in the event branding and activity slugs, and the event is described as a place to work through what finance leaders are actually doing with AI, where it creates real leverage and where it does not.
Epstein says the format responds to the limits of vendor demos and LinkedIn posts: finance leaders need to sit down with each other.
The tone is informal and direct, with lines such as "let's go to the moon with finance" alongside the numbers.
Repeated event-promotion structure: a short argument, the event and date, and a direct application link.
Two posts name a co-speaker and a partner brand in the event context. The material does not state any commercial terms, so ask directly about payment rather than inferring it from event branding.
For brands planning a campaign
Epstein's demonstrated production format is convening finance leaders in a room, not publishing standalone analysis at scale.
A webinar or executive roundtable collaboration matches existing practice and the audience they already assemble around AI in finance.
Ask for sponsorship arrangements and fees; the posts show event promotion and registration mechanics but do not state commercial terms.
The posts name two employers directly: Jeff Epstein writes that they were CFO at Oracle and are now an investor at BVP.
Read more
All five promote Finance Exchange, the October 15 event in San Francisco that the posts attribute to Zenskar, and ask finance leaders to register or apply to join.
They do not state Jeff Epstein's commercial arrangement with Zenskar; ask whether it is paid or otherwise compensated.
Survey figures credited to Consero and the CFO Monday Brief are cited research inside the writing, not brand relationships.
Oracle
Employment
In a post arguing that AI spending has to be justified like any other technology investment, Jeff Epstein describes an earlier role in the past tense: "As CFO at Oracle and other companies, I was the one trying to prove that technology investments were worth what we paid for them."
Read more
The post gives no dates for the role; ask for their tenure.
Jeff Epstein states the current role: "Now, as an investor at BVP, I'm the one asking founders to do the same" — asking founders to show what technology spending returned.
Read more
No portfolio company or individual investment is named in their posts; ask for examples.
Every post promotes Finance Exchange, the October 15 event in San Francisco.
Read more
One names it as "Zenskar Finance Exchange in SF on Oct 15"; the others carry the Finance Exchange name and hashtag and link to registration or an application to join.
Jeff Epstein says they will be part of the conversation on stage and urges finance leaders to attend.
The posts do not state their commercial arrangement with Zenskar; ask whether this is a paid engagement, a speaking slot, or another arrangement.
Jeff Epstein's public writing makes a recurring, concise argument about AI in the finance function, mainly through posts built around Finance Exchange, an in-person gathering for finance leaders in San Francisco.
Read more
Start with the posts on measurement—the question of what actually changed after AI was deployed in forecasting, reporting, close, collections, or analysis—then read those on agent accountability and close and reporting speed.
Nearly every post is tied to the event, with an application link and a report of demand, so read the writing as framing for a live conversation among finance leaders rather than standalone analysis.
The through-line is practical: define what agents may decide, keep a person accountable, and treat faster reporting as worthwhile only if decisions also change.
AI in financeAgentic AI governanceClose and reportingCFO perspectiveFinance process automationIn-person roundtablesCFOs and finance leadersFinance teams
Ahikam Kaufman
Kaufman writes about finance AI agents through the same operating questions Epstein raises: what agents cost to run, what controls they need, and what a team commits to when it builds or buys.
Read more
One post works through token consumption, permissions, approvals, segregation of duties, audit trails, and model governance for finance teams building their own agents.
Another asks what it would take to change direction once AI is running financial processes, naming embedded workflows, financial logic locked into configuration, and accumulating audit history as hidden dependencies.
That aligns with Epstein's argument that boundaries and accountability should be defined before agents act.
Both write for finance leaders evaluating the technology, not a general audience.
Kaufman's vantage is building and selling software: the posts promote unlimited token usage and a Financial Data Graph in a named platform, organizing the argument around consumption economics and system-replacement risk.
Epstein writes as a former CFO who now questions founders as an investor, focusing on whether AI changed the business at all.
Their recurring format is a live event for peers rather than a product argument.
Arif covers the same ground as Epstein from inside finance work.
Read more
One post sets out where AI helps an accounting team—flagging bank reconciliation anomalies, handling the obvious share of AP/AR work by pattern matching, and drafting variance commentary—while insisting a person reviews and approves each item and writes the narrative.
Another argues that a new CFO's first 90 days are about fixing alignment rather than metrics: boards, sales, and revenue recognition all hold different numbers, and clean dashboards built on an untrusted system get ignored.
Both ask how much of finance automation genuinely changes and where human judgment still has to sit.
Arif writes as an accounting advisor, describing client teams and the ratio of producing numbers to explaining them, with concrete process examples for senior accountants. Epstein focuses on measurement and governance questions for CFO peers, delivered through event promotion rather than practice-level process detail.
Radojicic shares Epstein's focus on what finance teams can do with AI tools and where the limits are.
Read more
One post describes building a custom reporting app with Claude Code, Git and GitHub, and Cloudflare, then lays out the rules that make it trustworthy: code calculates while AI explains; line owners review and edit every comment; and the process includes security, review and approval, MFA, role-based access, and a full audit trail.
Another, about an FP&A team on 70-hour weeks, argues that the difference between good and world-class teams is the standard of the templates they maintain, not the software they buy.
The human-in-the-loop and control themes closely track Epstein's posts.
Radojicic is a practitioner-educator publishing step-by-step build guides, downloadable FP&A templates, and a large newsletter for finance professionals. Epstein's posts are shorter arguments tied to a roundtable, aimed at provoking comparison among finance leaders rather than teaching a build or providing a resource.
Jeff Epstein’s posts address finance leaders weighing AI in finance operations, including forecasting, close, collections, reporting, and controls.
Read more
The content suits a CFO-focused campaign, but doesn’t establish who follows or sees the account. If verified CFO or finance-team reach is required, ask for recent audience survey results or another documented source showing job roles and seniority; use platform audience data separately to assess age or geography.
The posts show Jeff Epstein appearing at the Zenskar Finance Exchange in San Francisco on 15 October, a finance-leader event they promote alongside co-presenter Priya Rajan.
Read more
Access is by application or invite rather than open purchase: by early October, the organizers said the event was oversubscribed, and the posts point readers to apply to join or request an invite.
The audience is finance leaders — one post cites 200+ registered — and the stated aim is peer comparison rather than vendor pitches.
The posts contain no sponsorship policy, rate card, media kit, or partnership inquiry page; ask the organizers for those details.
Finance Exchange roundtable invitation
Buyer requirement
Access runs through an application or invite, not a published sponsorship slot.
The posts say the San Francisco event was oversubscribed by early October 2026, so a brand hoping to place its own finance leaders in the room should expect a selection step and limited seats.
Contact the event host about capacity before building an activation around attendance.
Co-presented AI-in-finance session
Buyer requirement
The sessions center on whether AI actually changed finance work — close, forecasting, collections, pricing, retention, controls — and explicitly contrast this with vendor demos.
A brand would need to agree on the topic and format with the event organizers and commit to an evidence-led agenda: what was spent on finance AI, what changed in the business, and how measurement works.
The posts show Epstein appearing alongside a named co-presenter, so plan for a shared stage rather than a solo slot.
Brief A, a B2B finance software company wants to reach CFOs evaluating AI in finance operations through a webinar or executive roundtable.
Read more
This is a strong fit for three reasons.
The subject matter matches the buyer's decision.
Epstein's posts focus on the questions a CFO asks before funding AI in finance: did the close, forecasting, collections, reporting or controls actually change; did margins, pricing or retention move; can any of it be measured.
A vendor selling into that evaluation does not have to teach them the category or steer them onto topic.
The format is already practiced.
All five posts promote one in-person finance-leader event, Finance Exchange in San Francisco, with an application or invite link and reported participation.
A roundtable or webinar is a collaboration they already run, rather than a format they would have to be talked into.
There is a ready-made angle.
The recurring frame, moving the question from "how are you using AI" to "what actually changed," gives a sponsor a session that is comparative rather than promotional.
A vendor willing to bring measured before-and-after results for named finance processes fits that frame; a vendor leading with capability claims does not.
Two conditions attach.
First, no post contains a sponsorship policy, rate card or media kit, so ask what paid partnership terms are available.
Second, the posts show a co-presented stage and invite-only access, so plan an activation as a shared session inside someone else's event, not a solo slot the brand controls.
When to choose a specialist
A consumer or small-business accounting app wants a keynote-style product webinar in which Epstein demonstrates the app, with guaranteed registration and attendance numbers and a paid media placement carried on the Finance Exchange brand.
Read more
This is a poor fit for specific reasons.
The content is comparison, not demonstration.
Epstein's posts repeatedly position peer discussion against product presentation, and one post says plainly that there is only so much to be learned from vendor demos.
A demo-led format asks them to do what their own material argues against.
The commercial mechanism is undocumented.
No post states a sponsorship policy, price, rate card, media kit or partnership inquiry route.
The posts don't support promising a paid placement inside the event, which is promoted by application or invite rather than as an open sponsorship slot.
Ask whether a paid partnership route exists and what terms are available.
Reach cannot be warranted from the posts.
They report 200-plus finance leaders registered and describe the event as oversubscribed, but provide no audience demographics, job-role breakdown or seniority data.
Ask what audience data the organizers can provide before committing to guaranteed impressions, named-account delivery or verified CFO composition.
The stage is shared.
A post names a co-presenter, Priya Rajan, so an activation assuming a solo keynote misreads the format before commercial terms are discussed.
A related misfit is a campaign aimed at non-finance audiences, such as developers or general consumers.
Every post addresses finance leaders and assumes fluency in close, collections, controls and pricing decisions.
A worked campaign example
Worked angle for a campaign aimed at CFOs evaluating AI-in-finance software.
Positioning.
Don't pitch a product webinar.
Pitch a measured-results session, with a working title along the lines of "What actually changed: finance leaders compare AI results."
The premise: AI is no longer a usage question, and the room should test outcomes process by process.
Why this angle holds.
Epstein already asks this publicly, listing forecasting, reporting, close, collections and analysis, and asking whether the team got more productive, whether the economics changed, whether decisions are faster or better, and whether any of it can be measured.
A sponsor that adopts the question instead of answering it with a capability slide works with the host's argument, not against it.
The vendor's credibility comes from being willing to be measured alongside the audience.
Session structure.
Open on the governance pressure point: a survey figure Epstein has cited is that 23% of finance leaders said that after a significant overnight AI agent error it would not be clear by 9 AM who was accountable.
Use that to frame the session as boundary-setting rather than autonomy-selling.
Then split into two blocks.
Block one: where agents may act unattended and what must escalate to a person, with the practical test of naming human ownership before naming an agent's permissions.
Block two: the close-speed problem.
A faster close is not a faster operating model if pricing decisions still wait for the monthly meeting, so ask each attendee which decisions would actually move if the reporting cadence did.
Close with the honest objection: finance teams report spending roughly a quarter of the week correcting AI output.
Ask the room what reduced it.
That admission makes the vendor's claims more credible to this audience, not less.
Sponsor role.
Offer working resources rather than a pitch: a one-page escalation checklist (who owns the decision, what an agent may do, what stops for a human) and a decision-cadence worksheet that maps reporting availability against actual decision timing.
Keep the product out of the session and put it in the follow-up, where finance leaders can evaluate it against their own numbers.
Format and tradeoffs.
An executive roundtable matches the practiced format better than a standalone webinar, but it is invite-limited and the organizers have described the event as oversubscribed, so plan for a selection step and small seats.
A webinar reaches more people but loses the peer-comparison character that makes this angle work.
A shared stage with a named co-presenter also means less brand control over the agenda; accept that, because the co-presented finance-leader format is the one the posts show.
Before committing budget, ask whether a paid partnership route exists and what audience data the organizers can provide on roles and seniority. Until the first is answered, the arrangement is unverified; until the second is answered, reach claims cannot be supported.
Choose Kaufman when a campaign needs someone who speaks from inside a finance-AI platform rather than from the buyer's chair.
Read more
Both posts take the vendor's side of the same question Epstein asks, and go deeper on operating economics than Epstein does: what a prompt, an agent action and a context window cost; how usage scales when autonomous agents run thousands of actions; and which model should run each process.
The second post pushes further into switching risk: whether deployment forces a system replacement, where the financial rules and process knowledge live, whether supporting documentation can be accessed independently, and what happens when an entity is added or an ERP changes.
It cites a Deloitte finding that 84% of finance leaders expect technology sovereignty to affect investment decisions through FY2027.
For a sponsor whose product is judged on integration, exit cost, permissions, approvals, segregation of duties, transaction-level traceability, audit trails and model governance, Kaufman covers ground Epstein's posts leave open.
The session resolves to a competitor.
Both posts close on the platform Kaufman helped build, named in the text as a way for finance teams to avoid managing token consumption, orchestration and controls themselves.
A sponsor would share a stage with someone whose answer to the problem is a different finance-AI product.
The format also does not match: neither post shows an event, roundtable or webinar, and neither reports participation numbers.
A shared finance-leader room is therefore not evidenced here as it is for Epstein.
The audience is unproven: there is no data on roles or seniority, and the writing addresses people weighing the buying decision rather than a demonstrated CFO room.
Choose Arif when the buyer's decision is about how AI changes the work of accounting and finance-operations teams, not how CFOs govern it.
Read more
The posts are concrete about where AI fits alongside people: in bank reconciliations, it flags anomalies while a human still reviews and approves each one; in AP and AR, pattern matching handles the obvious majority while the team keeps the messy remainder that needs judgment; in variance commentary, a machine-assisted first draft is refined by someone who knows the client's business.
The stated goal is to flip the ratio most finance teams live with — roughly 80% of time producing numbers and 20% explaining them — so senior people can spend their hours on why revenue fell in a segment, what cash looks like in 90 days, and whether the pricing model still works.
A second post covers a CFO's first 90 days, arguing that the real first task is alignment and trust in the numbers, not dashboards: clean dashboards that nobody believes get ignored.
That is a useful angle for a sponsor selling to finance teams mid-transition or selling onboarding and adoption rather than pure AI capability.
This is a practitioner-education voice, not a peer-comparison host.
Neither post carries governance, accountability or measurement material of the kind Epstein's posts build their argument on.
Neither names a co-presenter or event, gives a registration count, or establishes a room of finance leaders.
A brief promising reach into senior CFO decision-makers, or a session comparing AI outcomes across companies, would rest on evidence this material does not contain.
The team-ratio and coworker framing lands closer to controllers, senior accountants and finance-operations leads than to the budget-holding CFO evaluating a platform.
A sponsor targeting that CFO should expect to add seniority through another route.
Choose Radojicic when a campaign targets finance teams that build rather than buy, or the technical half of the finance-IT partnership.
Read more
One post follows a finance leader and small team building a custom reporting app with Claude Code, a Git and GitHub workflow, and a secure login.
They handle month-end work that falls between existing systems: pulling trial balances, budgets and forecasts; building management-format statements; comparing actuals against budget, forecast and prior year; and drafting variance commentary.
The post also lays out control rules a large vendor proposal usually skips: code calculates and AI explains; humans review and edit every comment before it reaches the CFO; MFA and role-based access; a full audit trail; and a developer review of the code and infrastructure before real data goes in.
The stated timeline is 10-to-14 weeks.
The other post audits an FP&A team on 70-hour weeks, with broken spreadsheets and three-day variance comments, and ends with a set of FP&A templates.
For a sponsor selling enablement, templates, tooling or training alongside software, this is closer to the audience's actual work than a CFO-governance conversation.
The posts describe owned media as the distribution route.
The session thesis points away from a packaged product.
It recommends a team build its own app around its own rules, so a vendor selling an off-the-shelf AI finance platform would be arguing with the room rather than supporting it.
The audience skews toward FP&A analysts and team leads who do the building, not necessarily the CFO who signs; a brief that depends on seniority or budget authority needs that confirmed elsewhere.
The format also does not suit a roundtable: the visible offer is a newsletter and downloadable template pack, a reach-and-nurture channel rather than an in-person executive event.
Neither post names a co-presenter or reports attendance, and no sponsorship or partnership route appears.
Jeff Epstein's five public LinkedIn posts here are complete post texts, not headlines or snippets.
They cover the same recent stretch of writing and repeat a small set of subjects: whether AI produced measurable change in forecasting, reporting, close, collections or analysis; who is accountable when an AI agent acts on cash, close or reporting; whether a faster close actually changes how the business decides; the share of a finance team's week spent checking AI output; and promotion of Finance Exchange, an October 15 gathering for finance leaders in San Francisco.
Two employer relationships are stated in the posts themselves — CFO at Oracle, previously, and investor at BVP now.
For comparison of subject and format, two posts each by Ahikam Kaufman, Anooshay Arif and Bojan Radojicic are included.
The posts do not include publication dates, so the writing cannot be placed in sequence and no claim is made about how the ideas developed over time.
These are five recent LinkedIn posts, not lifetime coverage: no book, show, article-length work, audience demographics, or sponsorship terms are included here, and any judgment about reach, paid partnerships or event ownership should be revisited if more information becomes available.
