Pharma ERP software for a Gujarat API manufacturer: manufacturing, exports and field sales in one system

What does GM Healthcare do, and why does it need two channels in one ERP?

GM Healthcare's actual business is manufacturing and exporting Active Pharmaceutical Ingredients from its plant at Saykha GIDC, Bharuch. The company's published profile covers a tagline, vision, mission, goals, a 2020 to 2025 milestone timeline, facility-design points and three named directors, and all of it appears on a Company Profile screen in the prototype.

The prototype models the API business and a branded formulation range side by side, because a pharma company that makes its own molecules often also sells finished products through chemists, distributors and agencies. The two channels behave very differently, so the ERP keeps them apart while sharing the plant.

The 11 API molecules are Benfotiamine, Carbamazepine, Minoxidil, Tadalafil, Dapoxetine, Oxcarbazepine, Etoricoxib, Hydrochlorothiazide, Quetiapine Fumarate, Hydroquinone and Linagliptin. Each carries its therapeutic use, regulatory filings, HSN code, per-kilogram contract pricing and its own batch inventory.

Backward integration is modelled. Four demo brands (Tadaforce 20, Dapoforce 60, Tadaforce-D and Benfogem 150) are formulated from GM Healthcare's own API. Open a brand and the product page shows the molecule it is made from; open the molecule and it lists the brands consuming it. That link is the commercial argument for making the API in-house.

Why custom pharma ERP software instead of a generic distribution system?

Generic ERP and billing tools treat stock as a quantity per product. Pharma needs it per batch, with an expiry date, a QC status and a licence-controlled customer at the other end. A pharma ERP also has to serve two commercial models, API export and formulation distribution, without mixing their units, terms and owners. A custom build lets the data model follow the plant, the field team and the export desk. That is the principle behind every AI-powered business management system Yukti AI builds, and it applies the same approach as its manufacturing ERP software.

The design rule was simple: there is no second set of dashboard numbers. Company total, territory performance, product ranking, customer ranking, leaderboard, incentive slabs and scheme progress are all roll-ups of the same order lines.

Pharma sales and distribution software: orders, customers and credit control

The sales side covers the order-to-cash view for chemists, distributors, agencies and hospitals, the counter types a pharma company in India sells to.

• Sales overview and orders. Order register with pending orders, order detail with batch allocation, and invoices that carry HSN-wise GST and an e-invoice IRN. The demo dataset holds about 1,470 orders and 1,400 invoices, generated from a fixed seed.
• Customer master. 432 demo accounts across 10 customer types, with pages for distributors, agencies, medical stores or chemists, and hospitals or institutions. Each customer page shows orders, ledger, schemes and the products they buy.
• Customer, product and territory analytics. Intelligence views for what a counter buys, which territories lead, and which customers are going quiet.
• Credit control. Sanctioned limit, utilisation and available credit per customer. A hard block applies at 100% with a logged override request. The field app runs the same check before an order is placed.
• Returns and credit notes. Sales returns with saleable, damaged and expired routes, and a credit-note and debit-note workflow.
• PTR and PTS price control. Separate retailer and stockist pricing, with MRP, PTR and PTS held per SKU and margin guardrails.

Pharma sales scheme software and a target engine with a pace line

Schemes and targets are where pharma sales teams lose the most time to spreadsheets. The prototype treats them as engines, not tables.

• Scheme engine, 14 mechanics. Buy X Get Y, quantity slabs, revenue slabs, cashback, free goods, gifts, travel rewards, performance schemes, product schemes, category schemes, new-customer schemes and collection incentives. Slab tables show the reward per tier.
• Liability only on qualification. Scheme cost accrues only when a customer or executive actually qualifies, so management sees real exposure, not a theoretical maximum.
• Scheme dashboard and campaigns. Separate views for customer, agency and medical schemes and campaigns. A demo campaign shows a closing-week leaderboard of participants who are qualified, almost there or behind. That campaign's cohort is calibrated for demonstration.
• Target engine. 7 target types, 4 periods and 4 assignee levels, measured against a pace line (the share of the month elapsed) rather than a flat percentage. It shows required daily run-rate and projected close.
• Incentives and payouts. Staff incentive slabs and incentive payouts computed from the same order lines.

Pharmaceutical manufacturing software: production, QC and purchase

For an API manufacturer, the plant is the product. These modules connect the shop floor to stock and sales.

• Production planning and orders. Production orders with stages, batch production records and planning against demand and stock.
• Raw material consumption. Consumption recorded per batch with variance against the standard.
• Quality check. QC records with test results and a release decision. Only released batches are eligible for FEFO allocation, and an alert escalates a batch that sits at QC for more than 48 hours.
• Suppliers, purchase orders and GRN. Supplier master, purchase orders, goods receipt notes, purchase invoices and purchase returns. Raw materials span API, excipient, packaging and solvent types.
• Reorder alerts. Raw-material and finished-goods reorder levels raise an alert and a draft purchase alert.

Collections and receivables: who owes what, and for how long

Accounts sees the same numbers the sales team sees. A typical path is receivables, then an ageing bucket such as 60+ days, then the customer, then the ledger, then record payment.

• Receivables and ageing. Outstanding invoices in ageing buckets, each clickable down to the customer.
• Customer ledger and outstanding. Full ledger per customer with statements.
• Collections and payments. Record receipts against invoices; collection incentives feed the scheme engine.
• Payables. Supplier payables from purchase invoices.
• Credit limits and expenses. Limits and utilisation, plus an expenses view.
• Margin reports. Profit and margin reports from the same order lines.

Pharma compliance software features built into billing and stock

Statutory details are part of the data model rather than a separate checklist.

Recall management, distributor and customer portals and secondary-sales capture are listed on a Pharma Capabilities screen as Beta or Planned, so the roadmap is stated openly instead of implied.

Every feature in the pharma ERP, and what it means for the business

A feature list on its own does not tell an owner much. What matters is what changes on an ordinary working day once the feature exists. The table below goes module by module through what Yukti AI built into the GM Healthcare prototype and the practical business benefit each one is designed to give. These are benefits the system enables; because the prototype runs on demo data, none of them is presented as a measured result.

How this system makes a pharma business AI-enabled, honestly

"AI" is used loosely in a lot of software marketing. This section separates what is genuinely in the GM Healthcare prototype today from what could sensibly be added next. The short version: the intelligence that exists today is data-driven and rule-based. It reads the ERP's own numbers, spots what needs attention and tells the right person. It does not call a generative AI model.

What is built today

• Ask Your ERP. An owner or manager types a question in ordinary language. The system recognises the products, customers, territories and executives named in it, works out the answer from the live roll-ups and links to the screen the number came from. In the code it covers questions such as an executive's orders, collections and rank; who is below target or behind pace; the sales leaderboard; stock of a named product by batch; batches nearing expiry; customers with the highest outstanding; collections; orders pending dispatch; which schemes cost the most; and a molecule's regulatory filings and HSN. Because it reads the same data as the dashboard, its answer always matches the screen.
• Data-driven alerts. The Alert Centre ranks problems by severity: a batch close to expiry with a suggested action, an executive falling behind the pace line, an account over its credit limit, stock below reorder level, a batch stuck at QC. Nobody has to go looking for these; the system raises them.
• Insight from history. The suggested order on the medical rep app is computed from each customer's purchase history, and the expiry action list ranks counters by how much of that product they have bought. Both are decisions made from data rather than memory.
• Explainable decisions. FEFO allocation does not just pick a batch; it states why. That matters in pharma, where a warehouse team and an auditor both need to understand the reasoning.
• Automation rules. Ten trigger-action rules turn insight into action: a reorder breach drafts a purchase requisition, a credit breach holds billing, a scheme slab locks the reward against liability.

Possible next steps (not built yet)

The following are natural extensions on the same data. They are not part of the current prototype, and are listed so the roadmap is clear rather than implied.

• A generative AI daily briefing that reads the roll-ups each morning and writes the owner a short note: what moved, which batches need pushing, which executives are behind and which customers are overdue.
• Demand forecasting by product and territory from order history, to guide production planning and raw-material purchase.
• Smarter reorder and scheme suggestions, for example proposing which scheme mechanic tends to move a slow product, based on past scheme performance.
• A WhatsApp ordering assistant for chemists, built on the WhatsApp Business API, that takes a repeat order in chat and creates it in the ERP. Yukti AI already builds WhatsApp AI chatbots and AI agents of this kind for other businesses.
• Collection risk flags that highlight customers whose payment pattern is slipping before they cross the credit limit.

The reason these are realistic next steps is that the hard part, one clean and consistent pharma dataset, already exists. AI is only as good as the data it reads, and in this design every number reconciles. That is the foundation of an AI ERP.

What a working day looks like in a pharma company running on one ERP

This is an illustrative day, built from the workflows the prototype supports. It shows how the different roles touch the same data without passing spreadsheets to each other.

Morning: the owner and the sales manager

The director opens the command centre and sees month-to-date sales against target, the leaderboard and total outstanding. The Alert Centre has already flagged a handful of batches inside the 60-day expiry window, two executives behind the pace line and one distributor over its credit limit. Rather than calling a meeting to collect numbers, the director asks Ask Your ERP which salespeople are below target and opens the staff 360 profile for the one furthest behind. The sales manager, looking at the same screens, adjusts that executive's beat for the day.

Mid-morning: the field force

Across the territories, medical representatives open the mobile app on today's visits. At each chemist they check in, start from the suggested order, let the scheme apply, pass the credit check and place the order. Where the expiry radar has pushed a batch to their list, they offer it to the counters that already buy that product. Where a customer has an overdue invoice, they record the collection on the spot.

Afternoon: the warehouse and dispatch

The warehouse manager works from the pending dispatch queue. Each picking list already shows the FEFO batch and the reason for it. Packing, transporter, LR number and the e-way bill for consignments above 50,000 rupees are recorded on the dispatch board, and a dispatch update is queued for the customer on WhatsApp (a simulated send in the prototype).

Through the day: plant, QC and purchase

The production manager tracks production orders by stage and checks raw-material consumption variance. QC releases tested batches, and any batch waiting more than 48 hours is escalated. A raw material dropping below its reorder level raises an alert and a draft requisition for the purchase manager. The export desk updates drum-wise consignments and letter-of-credit receivables for API buyers on the same plant data.

Evening: accounts and the close of day

Accounts opens receivables by ageing bucket, works through the 60+ day customers, opens each ledger and records payments received. By the end of the day the dashboard, the target screen, the scheme progress and the receivables all reflect what actually happened, because every one of them is a roll-up of the same order lines.

Technology and the path to production

The prototype is a React and TypeScript single-page web app that generates its dataset in the browser from a fixed seed, so every reload is identical and every number reconciles. It runs in a browser on desktop or phone with nothing to install, and is deployed on Vercel as a live demo with an automated deployment pipeline.

To go live the build would add a database and backend, real authentication and role enforcement, data migration from the client's existing records, and integrations such as Tally, the WhatsApp Business API, and e-invoice and e-way bill providers. The screens and data model do not need to be redesigned for that, because they were already reviewed against the seeded data.

Terms used in this case study

API

Active Pharmaceutical Ingredient: the active molecule in a medicine, made in bulk by a manufacturer such as GM Healthcare.

FEFO

First-expiry-first-out: sell the batch that expires earliest, provided it is released and has enough shelf life left.

GRN

Goods receipt note, recording raw material received from a supplier.

PTR / PTS

Price to retailer and price to stockist, the trade prices used alongside MRP.

Beat plan

A recurring route of customer visits for a field executive.

IRN

Invoice reference number issued under India's GST e-invoicing system.