Three Mumbai cutting-tool manufacturers replaced lakhs in annual Zoho subscriptions with a custom multi-tenant ERP that fits how they actually run.
The largest engagement Simply Five Studio has run to date. Three sister manufacturing companies in Mumbai. Three brands. One shared operating philosophy and three meaningfully different workflows under the same roof. They were spending lakhs of rupees a year on Zoho products and getting back software that did not match how their business actually moved.
The brief was direct. Customisations they needed sat outside what Zoho would allow. Workarounds had accumulated into a patchwork. The cost of the patchwork, in subscription fees, manual reconciliation, and lost visibility, was eclipsing the cost of building something custom.
CFX is the system that replaced it.
Carbiforce, Metaldur, and Kawacut manufacture cutting tools for industrial buyers. Solid carbide endmills, drills, reamers, indexable inserts, custom profiles. The category is technical. Customers know what they need. Specifications drive the conversation, not catalogue browsing. Sales cycles move through quotation, sampling, qualification, and repeat orders that become long-running supply relationships.
Three brands sit under common ownership. Each runs a slightly different go to market. The product catalogue overlaps in places and diverges in others. Some customers buy from more than one brand. Inventory is segmented but manufacturing capacity is shared. The mailing infrastructure, the customer master, the sales team coverage map, and the reporting all sit on the same operating fabric.
The previous tooling was Zoho. The pieces, taken individually, were serviceable. Zoho CRM for leads and quotes. Zoho Inventory for stock. Zoho Mail for communication. Custom fields, custom layouts, custom modules, custom reports. The cost was substantial. The fit, less so.
The customisations the founders wanted were unavailable. Cross-brand reporting that respected each brand's data isolation while still rolling up for the group. A sales workflow that handled the technical specification back and forth between buyer engineer and seller engineer, with versioned quotation documents rather than overwritten ones. Inventory transactions that reflected how the actual shop floor moved metal, not how a generic SaaS imagined movement. An Android app for field sales that worked offline at a customer site without a signal.
The founders had reached the realisation that comes to every mid-market buyer eventually. The subscription cost was now larger than the build cost of something purpose-built. The friction cost on top of the subscription was larger still. The work to make the SaaS bend was indistinguishable from the work to build something better.
They came to Simply Five Studio with that frame already articulated. The conversation did not need to argue for custom over SaaS. It needed to confirm that custom could be delivered to enterprise standards on a realistic timeline, and that the resulting system would be more reliable, not less, than the SaaS it replaced.
CFX is a multi-tenant operational system designed for three brands sharing one operational backbone. The architecture treats each brand as a logical tenant inside one application. Customer data, transactional data, and brand identity stay isolated where they need to. Shared resources, including the product master, the inventory ledger, and the sales team, present a single working surface across brands.
The system spans four major modules.
Lead capture from web, referral, and outbound. Lead enrichment with specification data captured at the point of intake so the sales engineer does not start from a blank slate. Opportunity tracking through stages that match the actual sales cycle, not a generic SaaS stage map. Quotation generation with versioned PDFs, brand-correct templates per tenant, and a quote-to-order conversion that does not lose the technical specification in translation.
Where Zoho imposed its own quote object model, CFX models the quote as a durable specification document. Buyers can iterate on a quote across several rounds. The system retains each version as a first-class artifact, which the sales team can refer back to and which the customer can request without a manual archive lookup.
A general ledger for goods. Movements rather than snapshots. Receipts from production, transfers between brand inventories, allocations against open orders, dispatches with carrier-attached documents, returns and rework flow. The inventory module knows that a brand's customer-facing catalogue maps to a multi-brand factory output, and reconciles correctly.
Stock cards present the full movement history for any SKU in any brand. The reporting on inventory turnover, slow movers, and stock-out exposure runs at both per-brand and group level, which Zoho's structure could not support cleanly.
A first-party mail layer integrated into the operational record. Email threads attach to opportunities, quotes, orders, and customer accounts automatically. The sales engineer does not switch between an inbox and a CRM to figure out where a conversation stands. The customer master maintains the canonical email history without manual filing.
Sender authentication, deliverability monitoring, and template management sit inside the application. The mail system is a feature, not an integration, which means it answers the same audit and retention questions the rest of the data answers.
A dedicated Android application for the field sales team. It is not a mobile-responsive web view. It is a native application designed for the field condition, which means it works without a network connection, syncs on reconnection, and presents the data the sales engineer needs at a customer site without scrolling through irrelevant fields.
The Gemini AI integration sits at the spec-translation layer. A sales engineer at a customer factory can capture a request in natural language, attach a photograph of an existing part, and let the model extract the specification fields the quote system expects. The model does not write the quote. It does the structural translation between conversation and form, which is the slow, error-prone work that field sales has always done by hand on a clipboard.
The model output is reviewed by the engineer before it commits to the system, and every committed record carries the audit trail of which fields came from the AI extraction versus which were entered or corrected by hand. The model is a productivity layer, not a decision layer.
A field sales engineer walks into a customer site in Pune or Aurangabad with the Android app. The customer engineer describes a job. The sales engineer captures the conversation, attaches a sketch or a photo of a sample part, and creates an opportunity that lands in the relevant brand's CRM at the headquarters in Mumbai. The estimator picks up the opportunity, validates the specification, calls back if a clarification is needed, and pushes a quote document into the brand's quote pipeline. The customer receives a versioned PDF with the right brand and the right contact. They reply by email. The email lands inside the opportunity record, not in a parallel mail thread.
If the customer asks for a revision, the estimator updates the quote specification and a new version is generated. Both versions remain accessible. The customer's eventual purchase order is reconciled against the latest agreed version. The production floor receives the order with the specification attached. Inventory commits material to the job. The order ships with a carrier-attached document set, and the customer account ledger updates on confirmation of dispatch.
Reporting rolls up by brand and across the group. The owners look at group performance in the morning and at brand performance throughout the day. Cross-selling between brands becomes a managed activity rather than a friction-laden accident. The sales engineers carry a Mumbai-headquarters operating system in their pocket, and the headquarters team sees what is happening in the field with no delay.
The annual subscription line went to zero. The customisation backlog, previously a permanent rolling list of things Zoho could not do, went to zero. The number of conversations that required a manual reconciliation between systems went to zero.
In its place is a system that does what the business does. Where the business changes, the system can change with it, because the business owns the codebase, the database, and the deployment. The marginal cost of an additional brand under the same operating shell is small, which matters as the group considers acquisitions and product line additions.
What the founders bought was not software. It was the structural flexibility to keep changing how they operate without being held hostage by a vendor's roadmap. The financial argument was secondary, and was satisfied in months. The strategic argument is the one that will compound.
CFX runs on AWS. Next.js for the application surface. PostgreSQL for the operational record. The Android sales app is a native build that talks to the same backend API as the web surface. Gemini AI is integrated directly at the application layer, without a middleware layer between the firm and the provider.
The choice of AWS, rather than a Hetzner or Hostinger VPS that the firm defaults to for smaller engagements, is a function of scale. Three brands, multiple regions of customer activity, an Android client base, and an operational tempo that justifies the operational tooling AWS provides. The decision to keep the application monolithic, rather than split into microservices, is a function of team size and operational simplicity. A custom ERP for a buyer of this profile rewards a single well-organised codebase over a distributed architecture that adds latency to every change.
The data isolation between brands is enforced at the application and query layer, not by separate database instances. This keeps reporting fast and operational maintenance simple, and the model is transparent enough that any engineer reading the code can verify the isolation guarantees.
CFX is a live engagement. New modules ship on a regular cadence. The group's appetite for what the system can absorb continues to grow as the team internalises that the constraint is no longer the software. The roadmap currently focuses on deeper production planning, finer-grained material costing per job, and broader use of the spec-translation AI layer in inbound buyer enquiries.
The relationship is structured as a long-term technology partnership, which is the shape every enterprise build at Simply Five Studio aims to become. The system is owned by the client. The continued evolution is run together.
A highly custom ERP with AI capabilities to remove manual workflows from the operation. Modules for customer management, HR, online portal automation, and custom quotations.
A comprehensive ERP for an industrial manufacturer of custom anchor bolts, flanges, and heavy hardware. Quotes to work orders to invoicing, multi-vendor RFQ, vendor and customer dashboards.
AK Suppliers & Distributors run a WooCommerce storefront integrated with a custom ERP we have evolved alongside their business for the last four years. The current stack is modern and entirely self-hosted.
Engagements begin with a conversation. Not a pitch. A direct discussion about your operations and the visibility you are missing.
