Zoho One: the operating system for a two-brand fintech

RBP Finivis is a forex company licensed by the Reserve Bank of India (RBI), with two brands: MEGO Forex, which serves travellers, students and families, and RBP Finivis, which licenses its platform to other businesses as a white-label product. It also runs an internship programme that trains students as referral agents.

That’s three very different customer journeys inside one company. I set up a large part of their Zoho One environment, so each journey has its own structure without breaking the others — and so the data coming out of it can be trusted.

• Three journeys, one CRM. A retail forex order, an institution onboarding onto a white-label platform and an intern becoming an agent have different stages, fields and compliance steps — identity checks (KYC), documents, payments.
• Data that looked right but wasn’t. Some picklist values displayed one label while storing a different value underneath, so any report on lead status counted the wrong stage.
• Decisions blocking the build. Several set-up choices needed leadership sign-off before configuration could continue.

1. Brand-specific layouts. Retail, White-label and Internship layouts on Leads and Deals, each carrying only the fields and stages that journey needs.
2. Three pipelines, each with its own stages and probabilities (below).
3. Custom modules for the agent programme — Cohorts, Agents, Agent Commissions and Agent Payouts, linked so each commission record points to an agent and a deal. (How the programme works.)
4. Capture and compliance fields — channel and service interest, service-level (SLA) tracking, a record of whether a conversation was handled by the AI bot, a person or escalated, and a consent block that records channel, source, notice version and timestamp.
5. Website integration — managed the backend API integration between Zoho and the corporate website, so enquiries arrive in the CRM with their context.
6. Operational structure across Zoho One — project and work management, employee records, lead-capture forms and shared documents, alongside the CRM.

The most important fix wasn’t visible. Several lead-status values showed one label to users while storing a different value in the database — a lead could appear to be at one stage while reports counted it as another.

I audited the values, corrected the mismatches and verified them. Anything built on top from here — dashboards, automations, forecasts — starts from lead-status data that means what it says.

Configuration work kept stopping on questions only leadership could answer. Rather than raise them one by one, I wrote a single decision memo for the managing director: eight decisions, each with the options, a recommendation, the urgency and exactly what it was blocking.

Three were closed the same day, and the remaining five moved to a tracked list of open decisions. It’s a small document, but it’s the difference between a rollout that drifts and one that moves.

For repetitive configuration and audits I used an AI browser agent working to a written brief: staged tasks, an evidence log of what it did, and a hard stop before anything irreversible. It waited at checkpoints for my explicit “proceed” before saving or submitting anything.

That let me move faster on routine work while keeping a human decision on every change that mattered — the same standard I’d expect from anyone touching a production CRM.

All three journeys now have their own structure inside one Zoho org: separate layouts and pipelines, a data model that links each commission to an agent and a deal, and consent and response-time fields on every lead. Automation and reporting are the next layer — built on data the business can trust.

CRM architecture · data modelling · pipeline design · process design · data quality · compliance-aware operations · stakeholder decision-making · AI-assisted administration.