Finance Agents

Tailored from Month-End Closer (Anthropic)

Cut days-to-close 30–40% by orchestrating the close checklist and drafting recurring journal entries with a full audit trail.

The period close is manual, serial, and reconciliation-heavy — inflating days-to-close and delaying reporting. This agent runs the close calendar, drafts recurring/accrual entries, and surfaces blockers so the team works exceptions, not the whole checklist. Moves days-to-close and manual-journal %.

Tailored from General Ledger Reconciler (Anthropic)

Cut intercompany reconciliation cycle time and intercompany mismatch rate by auto-matching transactions across entities and drafting correcting entries with a full audit trail.

Intercompany transactions across entities routinely fail to match on timing, FX, or missing counterparty postings, producing elimination breaks at consolidation that entity controllers chase manually. This agent auto-matches IC transactions across entities and currencies, ages unmatched breaks by root cause, and drafts the correcting and elimination entries. Moves intercompany mismatch rate and consolidation close time.

Tailored from Statement Auditor (Anthropic)

Cut the journal entry exception and SoD-violation rate by continuously screening every posted entry against SOX control rules in real time.

Manual journal entries are a leading source of financial-statement risk — SoD violations, missing support, and round-dollar or late postings slip through when review is sample-based and after the fact. This agent screens every JE at time of posting against a control rule library, flags exceptions in real time, and builds the audit evidence trail. Moves JE exception rate and control-testing effort.

Tailored from FP&A Analytics (KoGentic leading practice)

Cut variance-analysis cycle time 30–40% by auto-computing and decomposing actual-vs-budget variances into named drivers for every material line.

Analysts spend days each close manually pulling drivers for every material variance to build the flux narrative — a repetitive, data-intensive task that delays the reporting package. This agent computes variances, decomposes them into volume/price/mix/FX/one-time drivers, and drafts the narrative for analyst review. Moves variance-analysis cycle time and reporting-package delivery date.

Tailored from General Ledger Reconciler (Anthropic)

Lift balance sheet reconciliation on-time completion rate and shrink aged reconciling items by auto-matching GL balances to supporting schedules and drafting adjusting entries.

Balance sheet accounts are reconciled manually on spreadsheets, account by account, with aged reconciling items surfacing only at audit. This agent auto-reconciles GL account balances to supporting sub-ledgers and schedules, ages open items, and drafts adjusting entries for reviewer sign-off. Moves reconciliation on-time rate and aged reconciling item count.

Tailored from Statement Auditor (Anthropic)

Cut tax provision cycle time and shrink the return-to-provision true-up variance by auto-assembling book-tax workpapers and the effective-rate reconciliation each period.

The quarterly and annual tax provision is a manual, spreadsheet-heavy process pulling book-tax differences, deferred tax roll-forwards, and rate reconciliation from multiple systems under a tight close deadline. This agent assembles the provision workpapers, computes current/deferred tax and the effective-rate reconciliation, and flags return-to-provision true-ups for review. Moves tax provision cycle time and true-up variance.

Tailored from Model Builder (Anthropic)

Cut board and management reporting-package cycle time and lift on-time delivery rate by auto-assembling financials, KPI trends, and variance commentary into the board-ready template.

The board and management reporting pack is assembled by hand each period — pulling financials, KPIs, and commentary from multiple systems into a deck under a tight deadline. This agent assembles the financial statements, KPI trend charts, and variance commentary into a formatted board-ready package and routes it for executive review. Moves reporting-package cycle time and on-time delivery rate.

Tailored from Statement Auditor (Anthropic)

Shrink the ASC 842/IFRS 16 remeasurement backlog and lease/fixed-asset reconciliation breaks by detecting triggers and drafting remeasurement entries automatically.

Lease and fixed-asset accounting — ASC 842/IFRS 16 schedules, depreciation, impairment triggers, remeasurements — is tracked in spreadsheets alongside a dedicated lease system, and modifications or renewals often lag their required remeasurement. This agent reconciles the lease and fixed-asset registers to the GL, flags remeasurement/impairment triggers, and drafts the related journal entries. Moves lease-schedule accuracy and remeasurement backlog.

Tailored from FP&A Analytics (KoGentic leading practice)

Cut variance-analysis cycle time 50% by auto-explaining budget-vs-actual drivers each period with cited evidence.

FP&A burns days manually assembling BvA decks and hunting variance drivers. This agent pulls actuals vs plan, decomposes variances by driver, and drafts commentary — so analysts review, not assemble. Moves cycle time and forecast accuracy.

Tailored from Model Builder (Anthropic)

Improve rolling forecast accuracy 20% while cutting the forecast refresh cycle from weeks to days.

Rolling forecasts go stale because rolling actuals into the model and recalibrating drivers is a manual, spreadsheet-heavy chore each cycle. This agent auto-refreshes the forecast baseline off the latest actuals and driver trends and packages it for review — so FP&A spends its time judging assumptions, not rebuilding the model.

Tailored from Model Builder (Anthropic)

Cut scenario turnaround time from days to hours while widening board-ready coverage across best/base/worst cases.

Leadership asks for what-if scenarios arrive faster than spreadsheet-based FP&A can turn them around, and ad hoc models drift from the plan's assumptions. This agent builds and stress-tests best/base/worst and custom what-if scenarios off the current plan and quantifies the impact — so FP&A can answer same-day with a consistent, auditable basis.

Tailored from Model Builder (Anthropic)

Extend cash visibility to a rolling 13-week horizon while cutting cash forecast variance 30%.

Cash forecasting typically runs disconnected from the P&L/balance sheet plan, reconciled by hand against treasury and AP/AR feeds, leaving a short visibility horizon. This agent builds an integrated rolling cash flow forecast tying P&L, balance sheet, and working capital drivers together and reconciles it against actual bank positions — so treasury and FP&A share one forecast instead of two.

Tailored from FP&A Analytics (KoGentic leading practice)

Cut time-to-detect KPI breaches from weeks to hours by continuously monitoring finance KPIs against thresholds.

KPI breaches — margin erosion, DSO creep, opex overruns — are typically discovered only at month-end close review, weeks after they start. This agent continuously monitors a defined set of finance KPIs against thresholds and trend bands and alerts the accountable owner the moment a breach occurs, with root-cause context, so leadership acts within days, not weeks.

Tailored from FP&A Analytics (KoGentic leading practice)

Cut CapEx budget overrun rate and lift capital efficiency by flagging spend and schedule drift within days of occurrence.

Capital programs are typically tracked in disconnected spreadsheets and project systems, so overruns and schedule slips surface late and portfolio-level capital efficiency is hard to see. This agent monitors CapEx program spend against approved budget and milestones, flags overrun and slippage early, and rolls up portfolio capital efficiency — so FP&A and the PMO can redirect spend before it compounds.

Tailored from Collections & Cash Application (KoGentic leading practice)

Reduce DSO 15–20% by prioritizing collections outreach and drafting dunning at the right moment, within credit policy.

Collectors chase alphabetically, not by risk or value, and dunning is manual. This agent scores accounts by risk × value, drafts policy-bound outreach, and escalates disputes — lifting collections effectiveness and cutting DSO.

Tailored from Collections & Cash Application (KoGentic leading practice)

Lift cash application match rate to 90%+ by auto-matching remittances to open invoices and posting cash without manual touch.

Incoming payments arrive as lockbox scans, ACH files, wires, and checks with inconsistent or missing remittance detail, so AR analysts key matches invoice by invoice. This agent parses remittance data, matches payments to open invoices deterministically and probabilistically, auto-posts high-confidence matches, and routes the rest with suggested matches — cutting unapplied cash and speeding DSO recognition.

Tailored from Collections & Cash Application (KoGentic leading practice)

Cut dispute and deduction cycle time 30–40% by triaging, gathering evidence, and resolving deduction root causes within policy.

Disputes and deductions (pricing, shipping shortage, damaged goods, promo) sit unresolved for weeks because they route manually across sales, logistics, and finance with no single owner. This agent classifies the deduction reason code, gathers proof-of-delivery/pricing evidence, and either recommends resolution within policy or routes to the right owner with an SLA-tracked case.

Tailored from Collections & Cash Application (KoGentic leading practice)

Shrink DSO by generating accurate, contract-compliant invoices the same day an order or delivery is confirmed.

Invoices are drafted manually from order, contract, and delivery data, so pricing errors and late billing delay recognized revenue and irritate customers. This agent assembles order, contract, and delivery/usage data into an accurate draft invoice, validates it against contract terms and tax rules, and issues it on the customer's preferred format and channel.

Tailored from FP&A Analytics (KoGentic leading practice)

Improve the collections effectiveness index by forecasting 30/60/90-day cash collections within 5% of actuals from account-level payment behavior and risk signals.

Treasury and FP&A build cash forecasts from static aging buckets and gut-feel adjustments, so 30/60/90-day collection forecasts miss actuals and cash planning is reactive. This agent models payment behavior at the account level — combining historical payment patterns, promise-to-pay reliability, and macro/industry risk signals — to produce a rolling collections forecast with a confidence band, and flags cohorts drifting from plan.

Cut quote cycle time to under an hour for standard configurations by auto-generating compliant, correctly priced quotes through CPQ.

Reps manually configure and price quotes line by line, so straightforward, in-policy deals wait days for a quote while reps chase approvals for anything outside the standard price book. This agent auto-configures the quote from the opportunity and product rules, prices it against the approved price book and discount policy, and converts approved quotes straight to order — escalating anything outside policy for human approval.

Tailored from AP & Invoice Automation (KoGentic leading practice)

Lift touchless-invoice rate to 80%+ and cut duplicate/erroneous payments by validating every invoice against PO, receipt, and policy.

AP processes invoices manually with exceptions everywhere; duplicates and off-contract prices leak margin. This agent 3-way matches, checks tax/policy, and routes only true exceptions — lifting touchless rate and cutting leakage.

Tailored from AP & Invoice Automation (KoGentic leading practice)

Lift DPO by 3-5 days and capture 90%+ of available early-payment discounts without breaching liquidity or supplier-terms limits.

AP pays invoices on a fixed schedule that ignores discount windows and cash position, leaving early-pay discounts uncaptured and DPO below policy target. This agent sequences payment runs against discount terms, DPO targets, and daily cash position to release the highest-value run each cycle.

Tailored from AP & Invoice Automation (KoGentic leading practice)

Cut off-contract and no-PO (maverick) spend to under 5% of total procurement by enforcing PO policy at requisition and receipt.

Requisitions and receipts routinely bypass approved POs and contracted pricing, driving maverick spend and policy leakage that AP only catches after the fact. This agent checks every requisition and receipt against PO policy, contract terms, and approval limits in real time and blocks or routes non-compliant activity before it becomes a payment.

Tailored from Market Researcher (Anthropic)

Surface elevated supplier risk exposure early enough to act, holding at-risk concentrated spend under the board-set threshold.

Sourcing teams learn about supplier financial distress, ownership changes, or single-source concentration only after a disruption hits — by then it's too late to requalify or dual-source. This agent continuously monitors supplier financial health, news, and spend concentration and flags emerging risk before it becomes a supply disruption.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Cut demand forecast error 20-30% by blending the statistical baseline with real-time demand signals — so every downstream inventory buffer can shrink at the same service level.

Inventory buffers are sized to cover forecast error: every point of error forces safety stock the company pays to hold. Forecasts built on monthly history miss demand shifts that POS, open orders, and channel signals show weeks earlier. This agent senses demand continuously and maintains the accuracy-graded forecast every buffer downstream is sized from. Moves demand forecast accuracy, safety-stock dollars, and DIO.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Cut total inventory 15-25% at equal or better service by right-sizing safety stock and reorder points across the network to the finance-set working-capital target.

Safety stock accretes SKU by SKU and site by site — duplicated buffers, stale parameters, and single-echelon math leave the network holding more inventory than the target service level requires. This agent re-optimizes buffers multi-echelon (MEIO-style) against the finance-set service and working-capital targets, so the inventory policy is the minimum that meets service. Moves DIO, safety-stock dollars, and inventory turns.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Lift inventory turns and cut stockouts simultaneously by sizing, timing, and rebalancing replenishment to the optimized buffer policy.

Even with the right buffer policy, manual replenishment over-orders (round lots, gut-feel padding, set-and-forget order points) and lets stock sit in the wrong locations. This agent generates replenishment and rebalancing moves that hold the policy exactly — no padding — and drains overstock back through the network before new buys. Moves inventory turns, DIO, and fill rate.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Cut excess & obsolete inventory toward best-in-class levels by detecting aging stock early and driving disposition before it becomes a write-off.

Excess and obsolete stock is stranded cash: it inflates DIO, accrues carrying cost, and ends as a write-off nobody owns until the quarter it hits the P&L. This agent scans inventory aging continuously, quantifies the exposure in dollars, and drives a disposition decision (redeploy, markdown, liquidate, scrap) while recovery value remains. Moves E&O %, write-off dollars, and DIO.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Shrink the lead-time variability that forces buffer stock by tracking PO confirmations and inbound flow, and catching lead-time drift before planning parameters go stale.

Safety stock covers not just demand error but supply uncertainty: every day of lead-time variability forces buffer the company pays to hold. Planned lead times drift from reality quietly — suppliers slip, lanes slow — and nobody updates the parameters, so buffers are sized to fiction. This agent tracks confirmations, ASNs, and receipts against planned lead times, flags drift, and feeds lead-time reality back into buffer math. Moves lead-time variance, safety-stock dollars, and DIO.

Tailored from Inventory & Supply Planning (KoGentic leading practice)

Tie the demand-supply plan to the CFO's DIO and cash target every S&OP cycle — quantifying the working capital each plan option ties up or releases.

S&OP typically reconciles units and service; the cash consequence of the plan — the inventory it builds, the DIO it implies — arrives at the CFO after the fact. This agent sits between the planning cycle and finance: it translates every plan option into inventory dollars, DIO, and cash flow, runs service-vs-cash trade-off scenarios, and reports released working capital against the target. It is the finance-owned brain of the Forecast-to-Fulfill suite.

Give the CFO one live view of the whole finance agent fleet — every agent's SLA, exceptions, and cross-process handoffs — with escalations routed to the accountable human.

As the agent fleet grows, the operating risk moves from any single agent to the seams between them: a stalled close handoff, an unworked exception queue, an agent quietly failing its SLA. This supervisory agent watches the fleet itself — run status, exception aging, handoff completion across close → reporting → board — and routes what needs a human to the right owner. It coordinates; it never overrides another agent's guardrails.

Tailored from FP&A Analytics (KoGentic leading practice)

Release working capital to a CFO-set cash target by coordinating the DSO, DPO, and DIO levers as one cash-conversion-cycle program instead of three silos.

Receivables, payables, and inventory are each worked by their own team (and now their own agents) — but the CFO's number is the cycle: CCC = DSO + DIO − DPO. Optimizing silos leaves cash on the table and can even conflict (stretching DPO strains the suppliers whose lead times drive DIO). This agent reads all three lever families, computes the integrated cycle against target and peers, and sequences which lever to push next for the most cash at the least trade-off cost.

Tailored from Valuation Reviewer (Anthropic)

Widen the ROIC–WACC spread by ranking every capital deployment option — capex, M&A, R&D, buyback — on risk-adjusted return before the money is committed.

Capital allocation is the CFO's highest-leverage decision, yet most portfolios are assembled bottom-up from business-unit requests scored against inconsistent hurdle math. This agent normalizes every deployment option to the same risk-adjusted ROIC-vs-WACC basis, ranks the portfolio, and shows what the value-maximizing allocation looks like — so the committee debates trade-offs, not spreadsheets.

Tailored from Earnings Reviewer (Anthropic)

Close the say/do gap by reconciling the internal forecast to street consensus every cycle and pressure-testing guidance ranges before they're given.

Guidance credibility is an enterprise-value asset: a widening say/do gap costs multiple. Yet the internal forecast and street consensus are usually reconciled by hand, late, in the run-up to earnings. This agent maintains that reconciliation continuously — internal forecast vs consensus vs guidance — models the range the company can defend, and drafts the earnings Q&A from the gaps analysts will actually probe. Drafts only; everything external stays human.

Tailored from Model Builder (Anthropic)

Keep covenant headroom, FX exposure, and surplus-cash returns visible daily — so treasury acts on positions, not month-old reports.

The existing forecast agents project cash; nobody watches the capital-structure envelope around it — covenant headroom tightening, FX exposure drifting past hedge coverage, surplus cash idling below policy yield. This agent monitors positions daily against policy and flags what needs a treasury decision, with the numbers reconciled to bank and debt data. It never moves money.

Tailored from FP&A Analytics (KoGentic leading practice)

Recover margin leaking between list and pocket price by quantifying discount, rebate, and mix erosion at the transaction level and recommending price/mix actions.

Gross margin erodes in small, invisible steps — off-policy discounts, stacked rebates, freight give-aways, mix drift — that never appear in P&L summaries because each is individually immaterial. This agent walks the price waterfall from list to pocket at the transaction level, locates where margin leaks by product and customer, and recommends the price, discount-policy, and mix actions that recover it. Commercial owns every pricing decision.