SALESISell · Distribution & trade-spend intelligence

Catch the distributor that is slipping, and the claim that should not be paid, before the credit note goes out.

SalesI is distribution intelligence for companies that sell through distributors: FMCG first, with pharma, paints and agri packs. It scores distributor health, checks scheme claims against billing, returns and order patterns, and routes each claim to the right approver.

Run the demo Discuss your distribution challenge

Inside a small Indian grocery shop, shelves and hanging strips packed with packaged goods above sacks of rice

SalesI at a glance

1Reads

Distributor billing, payments and orders from DMS and ERP exports
Scheme rules, claims and claim lines
Returns, secondary sales, retailer orders and delivery records

2Decides

SalesI

Which distributors need a visit or a review this week
Pay, part-pay or hold each flagged claim
Where next quarter's trade budget goes

3Produces

Claims routed by amount and risk; low-risk claims up to ₹10,000 auto-approved
Credit note and payout recorded in SalesI after approval
Investigations opened on strong alerts

↺ Learns Outcome of each recommendation recorded, and its effectiveness tracked back to 1 · Reads

The business problem

A good distributor goes quiet before Diwali, then files its biggest claim

D-104 has billed every week for three years. In October its orders thin out, payments slip past terms and the rep misses two visits. Each fact sits in a different report, so nobody puts them together.
Then the Diwali scheme claim arrives: ₹1.8 lakh. Your billing for the scheme period shows ₹1.2 lakh. A third of last month's primary has come back as returns. The claims desk has hundreds of claims to clear before month-end and checks the paperwork, not the pattern.
The claim is paid, the stock sits in the market, and in January the distributor's secondary sales fall away. The warning and the leak were both in your data in October.

Why the usual fixes fall short

DMS and ERP reports show primary billing. They do not join billing, payments, visits, returns and claims for one distributor.
Claims desks check documents and approval limits. They rarely compare a claim with the billing and order pattern behind it.
Monthly distributor reviews catch the fall after the revenue has already gone.

Claims keep being paid on paperwork, and distributor risk keeps arriving as a revenue miss a quarter later.

Trade spendClaims above what was billed, and scheme targets met by stuffing, are paid out of margin.
RevenueA slipping distributor is noticed when sales fall, weeks after billing, payments and visits first changed.
Working capitalStock pushed on paper comes back as returns and credit notes, and receivables stretch.
Field timeReps follow the same beat every week, whatever is happening at the distributor.

In plain terms

What SalesI is, and how it fits

What it is: A distribution intelligence layer over your DMS and ERP data. It scores every distributor, checks claims and orders for leakage, routes claims for approval, and shows primary, secondary and retailer offtake side by side.
Who uses it: Heads of sales and distribution, finance controllers who own claims, trade marketing, and the ASMs, RSMs, reps and claims desk who work the queue every day.
What it does: Scores distributor health on five components, runs six leakage detectors (channel stuffing, fake and duplicate claims, ghost outlets, margin leakage, spike manipulation and fake secondary sales), routes claims by amount and risk, measures whether a scheme lifted sales, plans the rep's day, and allocates scarce stock to the markets that will sell it.
How it fits your systems: It reads Tally and Marg DMS exports, CSV or Excel uploads and SAP, read-only first. Connectors are configured per deployment. Credit notes are posted in your ERP by finance.
What is different: Every alert shows the rule that fired and an evidence score built from data completeness, freshness, rule checks and cross-source agreement. The scores are transparent rules and statistics you can read, and every claim above the auto-approve limit goes to a person.

See it work

SalesI in action

A distributor's health score falls 34% in two weeks, just as its Diwali claim arrives. Follow it from the warning to the claim decision and the recovery.

A simulated scenario on representative data. You choose the option and give the approval; the numbers follow your choices.

End-to-end workflow

From data to decision to outcome

What goes in, what SalesI works out, what gets decided, what happens next, and how the result comes back.

InputsDATA

Distributor billing, payments and orders from DMS and ERP exports
Scheme rules, claims and claim lines
Returns, secondary sales, retailer orders and delivery records
Rep visits and outlet lists with locations
Credit limits, receivables and marketplace listings

IntelligencePREDICT

Five-component distributor health score with tiers and an at-risk flag
Six leakage detectors, each alert with the rule that fired and an evidence score
Claim risk score: 40% history, 35% anomaly, 25% rule edge
Difference-in-differences scheme uplift and a trade-budget optimizer
Primary, secondary and offtake compared; outlets ranked and visits ordered

DecisionsDECIDE

Which distributors need a visit or a review this week
Pay, part-pay or hold each flagged claim
Where next quarter's trade budget goes
Which markets get scarce stock
Which outlets a rep visits first, and what to pitch

ExecutionEXECUTE

Claims routed by amount and risk; low-risk claims up to ₹10,000 auto-approved
Credit note and payout recorded in SalesI after approval
Investigations opened on strong alerts
Visits and pre-visit briefs added to the rep's day
Routine follow-up, such as payment reminders, by agents on the rung they have earned

FeedbackLEARN

Outcome of each recommendation recorded, and its effectiveness tracked
Health score history and tier changes
Investigation findings kept against each alert
Scheme and campaign uplift measured once they close
Agent actions and their ₹ value in the outcome ledger

Modules

Capabilities, organized by the work they do

28 modules in 5 groups. Open a module for its inputs, outputs, users and an example.

Distributors and field6 modules Schemes, claims and trade spend4 modules Leakage detectors7 modules Demand, commerce and insight7 modules Industry packs4 modules

Distributors and field

See which distributors and outlets are slipping, and send the field team where it changes the outcome.

Distributor command center and health scoreOne view per distributor, with a 0–100 health score on five weighted components and an at-risk flag when the score falls more than 15% in 14 days.

Challenge: A distributor slips for weeks across billing, payments and visits before the monthly report shows the revenue gap.
Inputs: Billing regularity, payment timeliness, order frequency, secondary-sales ratio and visit compliance, derived from DMS, ERP and visit data. Default weights 25/25/20/15/15, configurable. Sell-through by SKU for each distributor.
Outputs: Score history, a tier (platinum 85+, gold 70+, silver 55+, bronze below), an at-risk flag, the component that moved most, and which SKUs move and which sit at each distributor.
Used by: ASMs, RSMs, key-account managers, head of sales
Value: Sales teams work the at-risk distributors first, with the reason in front of them, instead of rebuilding the picture from spreadsheets and WhatsApp.
Works with: —
In action · illustrative: D-104 falls from 58 to 38 in 14 days: silver to bronze, at-risk flag on, payment timeliness the biggest drop. Two of its top ten SKUs have not moved to outlets in a month.

Credit and receivablesTracks each distributor's credit exposure, payments, outstanding and days sales outstanding.

Challenge: Credit keeps flowing to distributors whose payments are already slipping.
Inputs: Invoices, payments and credit limits.
Outputs: Credit limits, reconciled payments, payment reminders, collection tasks and DSO by distributor.
Used by: Credit and collections team, finance, ASMs
Value: Credit decisions use the same picture as sales decisions.
Works with: —
In action · illustrative: D-104's DSO moves from 32 to 51 days. A reminder goes out for the overdue invoices and a collection task is raised for the ASM.

Next-action recommendationsTurns signals such as a health score below 40, a falling scheme ROI or stock risk into ranked actions.

Challenge: Alerts pile up with no agreed response, so nothing happens.
Inputs: Signals from health scores, schemes and stock.
Outputs: Ranked recommendations, such as schedule a partner visit or start a health review, each with an evidence score, plus tracking of what happened next.
Used by: Sales operations, ASMs
Value: The response to each kind of signal is consistent, and its results are measured.
Works with: —
In action · illustrative: D-104 at 38 produces "schedule partner visit" first and "health review" second.

Field sales engineRanks a rep's outlets for the day, orders the visits, gives a brief before each visit and scores the visit against its objective afterwards.

Challenge: Reps visit by habit, so at-risk distributors and high-potential outlets wait, and nobody knows what to pitch on arrival.
Inputs: Outlet list with locations, revenue potential and depletion risk, PJP and visit-frequency rules, and the latest stock and scheme status for each outlet.
Outputs: A daily visit order (highest priority first, then the nearest next), a brief for each visit, geo-fenced check-in, audit templates and a post-visit score.
Used by: Field reps, ASMs
Value: Field time goes where it changes the outcome, and each visit is judged on what it achieved.
Works with: Check-in uses the phone's location; the geo-fence radius is set per company.
In action · illustrative: The ASM's Thursday beat opens with D-104, then the four outlets nearest to it. The brief for the first outlet: two fast SKUs below two days of cover, and no uptake of the Diwali display scheme.

Outlet churn riskScores each retail outlet on potential, activity and loyalty to flag those drifting away.

Challenge: Outlets stop ordering quietly, and nobody notices until the beat thins out.
Inputs: Outlet order and visit activity.
Outputs: A churn-risk score for each outlet, feeding beat plans and win-back campaigns.
Used by: ASMs, reps, trade marketing
Value: Retention effort reaches outlets before they are lost.
Works with: —
In action · illustrative: Twelve outlets on D-104's beat have not reordered in six weeks and are flagged for the ASM.

Retailer ordering and reorder nudgesLets retailers order on mobile or WhatsApp in their own language, with reorder suggestions from local sell-through and alerts on schemes and loyalty.

Challenge: Retailers wait for a rep to take an order, so reorders are missed between visits and scheme news reaches the shelf late.
Inputs: Retailer master and preferred language, local sell-through pace, the serving distributor's stock and price list, live schemes and loyalty rules.
Outputs: Retailer orders routed to the serving distributor, reorder nudges, scheme and loyalty alerts, and order history visible to the rep.
Used by: Retailers, distributors, field reps, trade marketing
Value: Reorders happen between visits, and the rep's time shifts from taking orders to selling.
Works with: WhatsApp Business messaging is configured per deployment, with opt-in and opt-out kept for each retailer.
In action · illustrative: A kirana in Nashik gets a WhatsApp nudge in Marathi: its 1 kg detergent is down to three days of cover at its usual pace. It confirms 12 units and D-104 receives the order.

Schemes, claims and trade spend

Write scheme rules once, check each claim before it is paid, and move trade budget toward what lifts sales.

Scheme builder and uptake trackingDefines schemes, who qualifies and the slab payouts, and tracks who took each scheme up and what it moved.

Challenge: Eligibility and slabs live in circulars and spreadsheets, so payouts are argued claim by claim.
Inputs: Scheme rules: tier, region, channel, minimum revenue or orders, and payout slabs.
Outputs: Enrolment, eligibility checks, tiered payouts, uptake and achievement tracking, and a leaderboard.
Used by: Trade marketing, sales
Value: Scheme rules are applied the same way to every distributor.
Works with: —
In action · illustrative: A festival scheme open only to gold and platinum distributors. D-104, now bronze, does not qualify for the next one, and the uptake view shows which eligible distributors have not enrolled.

Claim routing and credit notesRoutes each claim by amount and risk, and issues the credit note once it is approved.

Challenge: Every claim waits in one queue, so small clean claims are slow and large risky ones get the same light check.
Inputs: The claim and its lines, the scheme it is filed under, the distributor's claim history and anomaly signals.
Outputs: A risk score and a route: auto-approve up to ₹10,000 when risk is 0.30 or less, one approver up to ₹50,000, several above. Then a credit note and payout, with each approver recorded and a guard against settling the same claim twice.
Used by: Claims desk, RSMs, finance controller
Value: Clean claims move quickly, and risky ones reach the right people.
Works with: Limits are set per company. Credit notes are posted in your ERP by finance.
In action · illustrative: An ₹8,000 display claim with low risk is auto-approved. D-104's ₹1.8 lakh claim goes to the multi-level route.

Scheme ROI and budget allocationMeasures whether a scheme actually lifted sell-through, and proposes where the next trade budget should go.

Challenge: Schemes are renewed because they ran last year, not because they worked.
Inputs: Sales for scheme and comparison distributors before and after, scheme cost and budget limits.
Outputs: Uplift by difference-in-differences with a significance test, a cannibalization check, and a budget plan from a linear-programming optimizer.
Used by: Trade marketing, CFO
Value: Trade spend follows measured uplift.
Works with: —
In action · illustrative: The Diwali volume scheme shows no significant uplift in two regions. The optimizer moves part of that budget to a display scheme that did lift sales.

Channel marketing studioBuilds festival, win-back and scheme-uptake campaigns, sends them to outlets over WhatsApp with sector compliance rules applied, and measures them against sell-through.

Challenge: Trade marketing is manual and unmeasured: campaigns go out blind, and spend is never tied back to what moved.
Inputs: Audiences from outlet and distributor data (tier, churn risk, scheme uptake), campaign goal and budget, and sector rules such as pharma promotion limits.
Outputs: Campaign drafts with audience lists, compliance-checked messages, send schedules, and spend against incremental sell-through once the campaign closes.
Used by: Trade marketing, brand managers, compliance
Value: Campaign spend is judged on sell-through, the same way schemes are.
Works with: A language model drafts campaign copy; trade marketing approves every campaign before it is sent.
In action · illustrative: A win-back campaign targets the twelve outlets on D-104's beat that stopped reordering. Every message passes the compliance check, and four weeks later the report shows which outlets came back and what they bought.

Leakage detectors

Six checks that find money leaving through the channel, each with the rule that fired, the ₹ at risk and an evidence score.

Channel stuffingSpots primary billing pushed in to hit a target while secondary sell-through stays flat.

Challenge: Month-end stuffing meets the target, then comes back as returns and a false demand signal the next month.
Inputs: Primary billing, secondary sales and returns by distributor, with period-close dates.
Outputs: An alert when the secondary-to-primary ratio drops at month-end or returns pass the threshold (30% of primary by default), with the billing at risk of return in ₹.
Used by: Head of sales, finance, demand planners
Value: Targets met by stuffing are visible before the returns arrive, and planners stop reading them as demand.
Works with: —
In action · illustrative: D-104 billed a month-end spike in September, and returns came to 36% of primary against a 30% rule. ₹2.6 lakh of billing is flagged as at risk.

Fake and duplicate claimsChecks each scheme claim against actual billing and scheme eligibility, and finds the same claim filed under more than one program.

Challenge: Claims are paid on trust across thousands of lines, so inflated and duplicated claims slip through.
Inputs: Claims and claim lines, billed transactions from your ERP or DMS, scheme rules and slabs, and other open and settled claims.
Outputs: An alert when a claim exceeds billing by more than the threshold (20% by default), when the slab claimed is above the slab earned, or when a matching claim exists under another scheme.
Used by: Claims desk, finance controller, trade marketing
Value: The claims desk questions the right claims before the credit note goes out.
Works with: —
In action · illustrative: D-104 claims ₹1.8 lakh on 1,200 units for the Diwali scheme. Billing shows 800 units worth ₹1.2 lakh: 50% over the billing, and below the slab the claim was filed at.

Ghost outletsFinds outlets that order and draw scheme benefits but are never visited and never sell through.

Challenge: Fictitious outlets in the master quietly absorb stock and scheme money.
Inputs: Orders by outlet, rep visit records, DMS secondary sales and the outlet master.
Outputs: A ghost-outlet flag when an outlet has orders but no rep visit in 90 days and no sell-through behind them. A person can freeze its benefits and open an investigation.
Used by: Sales operations, finance, internal audit
Value: Stock and scheme money stop flowing to outlets that exist only in the master.
Works with: —
In action · illustrative: Four of the 20 outlets on a display claim have orders but no rep visit in 90 days. ₹8,800 of display payout is held for review.

Margin leakageCompares the discount given on each order with the discount the scheme allows, and flags off-scheme price cuts.

Challenge: Unauthorized discounts and stacked schemes erode margin one order at a time, and stay invisible until the quarter closes.
Inputs: Order lines with price and discount, price floors, and the schemes each distributor and outlet is eligible for.
Outputs: Alerts for prices below the floor and for discount stacking, with the ₹ gap on each order.
Used by: Finance, pricing, RSMs
Value: Margin is checked at order time instead of at quarter-end.
Works with: —
In action · illustrative: A 4% cash discount is stacked on a 6% volume scheme for an outlet eligible for the volume scheme only. Each order shows the 4% gap in rupees.

Spike manipulationFinds order spikes timed to incentive or scheme cut-offs that collapse the week after.

Challenge: Targets and incentive windows are gamed with artificial spikes, and sales fall away the following week.
Inputs: Order velocity by distributor and outlet, its normal baseline, and incentive and scheme cut-off dates.
Outputs: A flag before payout when orders in the window before a cut-off run well above baseline and fall back after, linked to the claim or incentive it would pay.
Used by: Claims desk, sales operations, HR for incentives
Value: Incentives pay for demand, not for timing.
Works with: —
In action · illustrative: D-104's orders ran 2.6 times normal in the five days before the Diwali deadline, then dropped below baseline. The flag sits on the claim before it is paid.

Fake secondary salesCross-checks secondary sales entered in the DMS against delivery records, to find sell-through with no movement behind it.

Challenge: Secondary entries are faked to qualify for schemes, and the whole sell-through view loses its value.
Inputs: DMS secondary invoices, and delivery and proof-of-delivery records from your dispatch or transport system.
Outputs: Mismatch flags for secondary entries with no delivery, the scheme benefit they would earn, and a clean secondary series for planning.
Used by: Sales operations, finance, demand planners
Value: The secondary view used for schemes and planning reflects stock that actually moved.
Works with: —
In action · illustrative: 140 cases billed as secondary to six outlets in the last three days of the month have no delivery record. The entries are flagged and left out of the clean sell-through series.

Leakage investigationsTurns a strong alert from any detector into a tracked investigation with an owner and findings.

Challenge: Suspicions are raised in email and go no further.
Inputs: Alerts and analyst findings.
Outputs: Investigation records with status and findings, bulk assign and dismiss, and a workflow that opens automatically when an alert's evidence score is 0.70 or higher.
Used by: Finance, internal audit
Value: Every alert ends with a finding: recovered, explained or dismissed.
Works with: —
In action · illustrative: The D-104 stuffing alert, with an evidence score of 0.8, opens a workflow assigned to the regional finance analyst.

Demand, commerce and insight

See real demand across every channel, put scarce stock where it will sell, and get answers and routine follow-up with the evidence attached.

Secondary sales and revenue viewPuts primary billing, secondary sell-through and retailer offtake side by side for each distributor, SKU and territory.

Challenge: Primary, secondary and offtake are seen in separate reports, so stuffing hides and planners plan on billing instead of demand.
Inputs: Primary billing from the ERP, secondary from the DMS, retailer offtake from retailer orders, and modern-trade and online feeds.
Outputs: A cross-channel view, an estimate of stock sitting in the channel, stuffing and pile-up flags passed to the leakage detectors, and a clean demand series for planning.
Used by: Head of sales, demand planners, finance
Value: Planning runs on demand that actually reached the shelf.
Works with: Feeds PlanningNest as a demand signal when both are in use.
In action · illustrative: Primary for detergents in Nashik is up 18% in September while secondary is flat and offtake is down 3%. The view shows about five weeks of stock sitting in the channel.

Allocation and available-to-promiseAllocates constrained factory stock to the markets most likely to sell it, and gives available-to-promise quantities and dates that supply can keep.

Challenge: Scarce stock goes to whoever pushes hardest, so the right product starts in the wrong place.
Inputs: Available stock and the production plan, open orders, distributor health, and sell-through and channel stock by market.
Outputs: An allocation proposal by market and distributor, ATP quantities and dates for each order, and the reason each market got what it got.
Used by: Sales operations, supply planners, RSMs
Value: Constrained stock goes where it will sell, and commitments to distributors are ones supply can keep.
Works with: Reads stock and plans from your ERP, or from Stocki and PlanningNest when in use. Sales operations approves each allocation before it is released.
In action · illustrative: A festival pack is 20,000 cases short. The proposal favors markets with thin channel stock and strong sell-through, and holds back markets already carrying five weeks of stock.

Territory intelligenceShows where the brand is strong and weak, postal code by postal code, finds uncovered clusters and forecasts festival peaks ahead of the season.

Challenge: Nobody can see weak coverage below district level, or see a festival surge early enough to stock for it.
Inputs: Sales and outlets with locations, territory boundaries, outlet potential, and festival and promotion calendars.
Outputs: Postal-code heat maps, white-space clusters ranked by potential, coverage gaps, and festival demand uplift by territory.
Used by: Head of sales, ASMs, trade marketing, planners
Value: Expansion and festival stocking follow the map instead of habit.
Works with: —
In action · illustrative: Thirty-eight postal codes around Nashik have high-potential outlets that no beat covers. They are ranked and handed to the ASM as a coverage plan.

E-commerce and quick-commerce intelligencePuts general trade, modern trade, marketplaces and quick commerce on one screen: price and promo parity, channel conflict and dark-store stock-outs.

Challenge: Online and offline teams work apart, so online prices undercut distributor trade and dark-store stock-outs go unseen.
Inputs: Marketplace and quick-commerce listings, prices and stock, MRP, the distributor's price to stockist and scheme prices, and product reviews.
Outputs: Parity alerts against MRP and trade prices, channel-conflict flags sent to the leakage queue, stock-outs by dark store, and themes from reviews.
Used by: E-commerce team, key accounts, trade marketing
Value: Trade trust holds because online undercutting is seen and handled, and dark-store gaps close sooner.
Works with: Marketplace and quick-commerce data arrives through seller APIs or file uploads, configured per deployment.
In action · illustrative: A 1 kg pack sells online for ₹8 less than D-104 pays after its scheme. The conflict goes to the key-account manager with the listing and the trade price side by side.

Sales copilot and weekly narrativesAnswers plain-language questions across the distribution chain and writes a weekly leadership summary, with every statement linked to its data.

Challenge: Leadership reads stale decks, and what moved, what is at risk and what to do is buried in spreadsheets.
Inputs: Health scores, claims, leakage alerts, scheme results, and sales and stock data.
Outputs: Answers with the figures and sources behind them, a weekly narrative (what moved, what is at risk, what to do), and draft actions that wait for confirmation.
Used by: Head of sales, CFO, regional leadership
Value: Leaders start the week from the same evidence the field sees.
Works with: A language model writes the prose through the AI gateway; figures come from the scoring and detector engines, never from the model. A proposed action lapses unless confirmed within five minutes.
In action · illustrative: "Which Nashik distributors dropped a tier this month, and why?" returns D-104 and two others, the component that fell for each, and the claims open against them.

Agentic revenue loopsAgents that take on routine follow-up, such as chasing overdue payments or missing data feeds, on an earned-autonomy ladder, with every action and its value written to an outcome ledger.

Challenge: Dashboards spot a problem and then wait for a person, and nobody lets an unaccountable AI near money.
Inputs: Signals from receivables, health scores, stock, schemes and data checks, and the autonomy rung set for each agent.
Outputs: Drafted or executed actions (collection reminders, data-chase requests, partner interventions, scheme and win-back proposals), an outcome ledger of ₹ recovered, leakage prevented and DSO change, and moves up or down the ladder.
Used by: Sales operations, credit and collections, finance
Value: Routine follow-up happens on time, and autonomy is given one task at a time on a measured record.
Works with: Rungs: draft and approve, automatic below a threshold, then automatic. An agent moves up only on a high approval record and moves down when bad outcomes pass a limit. Actions above ₹50,000 of impact need a person by default.
In action · illustrative: The collections agent drafts WhatsApp reminders for D-104's three overdue invoices. It is on the draft-and-approve rung, so the ASM sends them, and the payment that follows is logged against the action.

Distribution benchmark indexCompares your distributors with an anonymous sector cohort on measures such as OTIF, growth and DSO.

Challenge: Every distributor is graded against its own last month, so nobody knows what good looks like in the sector.
Inputs: Your distributor metrics, shared only if you opt in, and cohort definitions by sector and region.
Outputs: Your percentile against the cohort median on each measure, and where you trail most.
Used by: Head of sales, CFO, strategy
Value: Targets are set against the sector, not only against last year.
Works with: Opt-in and off by default. A cohort figure is shown only when at least five companies contribute, and raw values are never shared.
In action · illustrative: West-region distributors sit in the bottom third of the cohort on DSO but the top quarter on fill rate. Collections, not supply, is the gap to work on.

Industry packs

The same core, with the rules, signals and vocabulary each industry needs.

FMCG packAdds daily beat plans and PJP across a large outlet universe, festival demand forecasts and trade promotions measured on sell-through.

Challenge: Generic sales-force and DMS tools give no festival foresight, thin secondary visibility and beats that ignore revenue.
Inputs: Outlet universe, PJP rules, festival and promotion calendars, secondary sales and trade-promotion terms.
Outputs: Revenue-ranked beats and PJPs, festival demand uplift by territory and SKU ahead of the season, and promotion results on sell-through.
Used by: Sales, trade marketing, demand planners
Value: Festival stock and field time are in place before the peak.
Works with: —
In action · illustrative: Six weeks before Diwali, the festival model raises the gift-pack forecast for Nashik, and the beat plan adds a second weekly visit to the outlets that sold it last year.

Pharma packAdds drug-schedule rules and license tracking, medical-rep engagement with prescribers, and batch traceability with FEFO and recall.

Challenge: Pharma distribution carries legal rules that FMCG tools ignore: schedule classes, licenses, batch traceability and expiry.
Inputs: Product schedule classes, distributor and chemist drug licenses, MR visit and sampling records, and batch and expiry data.
Outputs: Order checks against schedule and license, license-expiry alerts, prescriber engagement views, batch traceability to the chemist, FEFO and recall lists, and a compliance audit log.
Used by: Distribution, regulatory affairs, sales-force effectiveness, quality
Value: Compliance is checked on every order, and a recall starts from a complete list.
Works with: —
In action · illustrative: A recall on batch B2241 lists the 41 chemists who received it through three distributors, with the quantity still on hand at each.

Paints and coatings packTreats construction projects as demand signals, maps contractor and painter networks, and manages tinting machines and shade trends at dealers.

Challenge: Paint demand comes from projects and contractors and is tinted at the dealer, which packaged-goods reorder tools cannot see.
Inputs: Project leads and stages, contractor and painter records and loyalty, tinting-machine placements and tint volumes by shade.
Outputs: Stage-based project demand, contractor and painter influence maps, loyalty status, and tinter use and shade trends by dealer.
Used by: Sales, dealer development, trade marketing
Value: Dealers are stocked for the projects and shades that are coming.
Works with: —
In action · illustrative: A 300-flat housing project reaches the finishing stage. Expected exterior emulsion demand rises for the two dealers nearest the site, and the lead contractor's painters are offered the loyalty scheme.

Agri inputs packAligns demand to the Kharif, Rabi and Zaid seasons and weather at taluk level, and tracks demo plots through to farmer purchases.

Challenge: Agri demand follows crops and rainfall, and a demo plot sells more than a rep's pitch.
Inputs: Crop calendar by taluk, weather data, dealer and retailer sales, demo-plot records and farmer leads.
Outputs: Season- and weather-adjusted demand by taluk and product, demo-plot conversion tracking, and dealer-to-farmer reach.
Used by: Sales, agronomy, dealer development
Value: Stock arrives for the sowing window, and demo plots are judged on purchases.
Works with: —
In action · illustrative: Early monsoon rain in two talukas brings the Kharif herbicide window forward. Dealer stock moves up two weeks, and farmers who visited the demo plot are followed up through their dealer.

Intelligence and automation

What the intelligence actually does

Five levels, from seeing what is happening to learning from it.

Visibility

Each distributor's health score, tier and the component that moved, next to its claims, returns, receivables, and primary, secondary and offtake.

Prediction

An at-risk flag when the score falls more than 15% in 14 days, outlet churn risk, festival demand by territory, and alerts on claims and order patterns that point to leakage.

Decision

Ranked next actions such as a visit or a review, a claim route by amount and risk, a trade-budget allocation, and a stock allocation for constrained products.

Execution

After approval, the credit note and payout are recorded in SalesI, visits go onto the rep's day, and agents carry out routine follow-up on the rung they have earned. Posting the credit note in your ERP stays with finance.

Learning

The outcome of each recommendation and agent action is recorded, so the effectiveness of each response is measured over time, and agents move up or down the autonomy ladder on their record.

Runs automatically

Health scores and tiers when new data arrives
Leakage checks on claims, orders and secondary entries
Auto-approval of claims up to ₹10,000 with a risk score of 0.30 or less (limits are yours to set)
Ranking and ordering of a rep's visits
Agent actions on a rung the agent has earned, below the limits you set

Always waits for a named person

Claims above the auto-approve limit: one approver up to ₹50,000, multi-level above
Holding, cutting or rejecting any part of a claim
Changes to scheme eligibility or slabs, and every campaign before it is sent
Moving trade budget between schemes, and allocating constrained stock
Agent actions above ₹50,000 of impact by default

Business value

What it helps you work toward, and how we measure it

We agree the baseline with you in writing before a pilot starts. These are the outcomes we measure.

Claims paid on evidence

Claims above billing, or tied to stuffing, are questioned before payment.

Flagged ₹ held, recovered or confirmed, by detector, against claims paid in the same period.

Earlier warning on distributors

The fall shows up in billing, payments and visits before it shows up in revenue.

Days between the at-risk flag and any revenue drop, and the share of at-risk distributors visited within a week.

Faster clean claims

Low-risk claims stop waiting behind risky ones.

Claim cycle time by route (auto, single approver, multi-level), before and after.

Trade spend that lifts sales

Budget moves toward schemes with measured uplift.

Scheme uplift by difference-in-differences, and budget moved away from schemes with no significant lift.

A demand signal you can plan on

Primary, secondary and offtake are read together, so stuffing stops passing for demand.

Gap between primary and secondary by distributor, and stuffing alerts closed, month on month.

Illustrative example

Illustrative: a ₹1.8 lakh Diwali claim against ₹1.2 lakh of billing. Paying the billed portion and holding ₹60,000 for investigation keeps the distributor whole on proven sales while the gap is checked.

Enterprise readiness

Integration, control and deployment

ERP & system integration

Reads Tally and Marg DMS exports, CSV and Excel uploads and SAP, read-only first. Marketplace and quick-commerce data arrives through seller APIs or file uploads. WhatsApp Business, email and SMS carry outbound messages. Connectors are configured per deployment, and credit notes are posted in your ERP by finance.

Data sources & connectivity

Works from your billing, payments, orders, returns, claims, visits and delivery records, joined into one view of primary, secondary and retailer offtake for each distributor.

Deployment & implementation

Services with Postgres, run as a managed service. The hosting region is agreed per deployment.

Configuration & business rules

Health-score weights, tier cut-offs, detector thresholds, claim routing limits, agent autonomy rungs and scheme rules are set per company. FMCG, pharma, paints and agri packs are switched on per company.

Roles, access & approvals

Roles for sales leadership, ASMs and reps, the claims desk, finance and trade marketing, with maker-checker on claims. Sign-in uses your identity provider.

Security, governance & audit

Tenant ID on every table with row-level security policies. No security certification is claimed.

Human oversight & exceptions

Claims route to people by amount and risk, and each approver is recorded. Every alert shows its rule and evidence score. Agents work on an earned-autonomy ladder, and each action is written to the outcome ledger.

Scale & ongoing optimization

Built for many companies and thousands of distributors on one platform, separated by tenant.

Straight answers

Questions about SalesI

Something else? Ask a specialist.

Does SalesI block claims by itself?

No. It flags and routes. By default, claims up to ₹10,000 with a risk score of 0.30 or less are auto-approved, and you set those limits. Anything flagged or larger goes to a person, and holding or cutting a claim is always a person's decision.

What goes into the distributor health score?

Five components: billing regularity, payment timeliness, order frequency, secondary-sales ratio and visit compliance, weighted 25/25/20/15/15 by default. You can change the weights. A fall of more than 15% in 14 days raises an at-risk flag.

Does it check every claim against its scheme slab?

Yes. Each claim is compared with the billing behind it, with the slab the scheme allows for the units actually billed, and with claims filed under other schemes. A claim that fails any check goes to a person with the reason.

Which systems does it read?

Tally and Marg DMS exports, CSV or Excel uploads, and SAP, read-only first. Marketplace and quick-commerce data comes through seller APIs or file uploads. Connectors are configured per deployment, and credit notes are posted in your ERP by finance.

Can the agents act without a person?

Only where you let them, one task at a time. Each agent starts on draft and approve, where a person sends every action. It moves to automatic below a threshold, and then to automatic, only on a strong approval record, and it moves back down when bad outcomes pass a limit. Anything above ₹50,000 of impact waits for a person by default.

Next step

Discuss your distribution challenge

SalesI goes after money you have already earned and are about to give away: claims above billing, scheme targets met by stuffing, distributors slipping unnoticed. Every flag shows its rule and its evidence, and a person decides.

Talk to a specialist

Manufacturers selling through distributors in India (FMCG first, also pharma, paints and agri) who pay scheme claims every month and want them checked before payment.

In the same loop