Work · 2023 → Now

Software companies actually run on.

Each project below is a production system: real architecture, real shipped features, a real codebase owned by the client. No mockup-only concepts and no inflated metrics — just the work, broken down honestly.

Custom-built

every system from scratch

Production-grade

deployed and in use

Client-owned

full codebase handover

Documented

architecture & runbooks

Featured Case Studies

Featured systems, end-to-end.

Click through for the full breakdown — challenge, solution, architecture, workflow and outcomes.

AI Revenue & Quotation Platform — AI lead qualification interface with conversation summary, requirements extraction, scoring and recommended next actions

In production

Interior Design · AI Sales OperationsRead case study

AI Revenue & Quotation Platform

Custom AI sales and quotation platform built for interior design and renovation companies — automating the entire journey from first enquiry to approved quotation.

System features

AI lead qualification
Conversation analysis
Project requirements extraction
Floor plan analysis
AI quotation engine
Lead scoring & prioritisation
+4 more

Stack: React · TanStack · Tailwind · Node.js
Architecture: Customer conversations, AI qualification, document analysis and quotation generation share a single operating system — designers and management work from the same source of truth.

Healthcare Scheduling & Patient Operations Platform — appointment booking and patient operations interface

In production

Healthcare · Clinic SoftwareRead case study

Healthcare Scheduling & Patient Operations Platform

Custom healthcare operations software built to streamline appointment scheduling, intake workflows, recurring services, approvals, patient management, and administrative operations.

System features

Appointment scheduling
Availability management
Multi-service booking workflows
Recurring appointment logic
Intake and onboarding forms
Administrative controls
+4 more

Stack: React · TanStack · Tailwind · Node.js
Architecture: Patient-facing app and clinical admin console share a single backend with role-based access, a Postgres source of truth and automated messaging on top.

Financial Modelling & Investment Simulation Platform — portfolio management dashboard with unrealised, realised and total P/L, open positions across multiple equities, watchlist, search and allocation analytics

In production

Fintech · Financial Services · EducationRead case study

Financial Modelling & Investment Simulation Platform

A custom financial modelling and investment simulation platform for investment training, portfolio analysis, scenario forecasting and market research — simulate market conditions, evaluate strategies and analyse portfolio performance in a single controlled environment.

System features

Portfolio management
Investment simulation
Scenario modelling
Market analytics
Trade tracking
Recommendation engine
+4 more

Stack: React · TanStack · Tailwind · Data visualisation
Architecture: Portfolio surfaces, simulation services and market data share a single financial operating environment — analysts, traders and learners work from the same source of truth.

Luxury Commerce & Operations Platform — unified commerce operations dashboard with revenue analytics, orders, customer management and inventory oversight

In production

Luxury Retail · Commerce & OperationsRead case study

Luxury Commerce & Operations Platform

Custom commerce operations platform built for a luxury direct-to-consumer retail brand.

System features

Order management
Payment verification workflows
Customer management
Product and inventory tracking
Revenue analytics
Abandoned cart recovery
+4 more

Stack: React · TanStack · Tailwind · Node.js
Architecture: Admin console and storefront share a unified commerce backend — orders, inventory, payments, customer records and marketing operations in one system.

Case 01 · Interior Design · AI Sales Operations

AI Revenue & Quotation Platform

Custom AI sales and quotation platform built for interior design and renovation companies — automating the entire journey from first enquiry to approved quotation.

In productionClient-owned codebase

Challenge

Interior design firms were spending significant time manually responding to enquiries, qualifying leads, reviewing floor plans, preparing quotations and following up with prospects. Response times varied, lead quality was inconsistent, and sales teams were spending too much time on administrative work instead of closing projects. The result: slow response times, lost revenue from missed follow-ups, poor visibility into the pipeline, and difficulty scaling without hiring more staff.

Solution

Designed and implemented a complete AI-powered revenue operating system that captures leads automatically, qualifies opportunities using AI, extracts project requirements, analyses uploaded floor plans, generates quotation recommendations, scores lead quality, prioritises sales opportunities, automates customer communication and gives management full visibility across the sales pipeline.

Shipped features

AI lead qualification
Conversation analysis
Project requirements extraction
Floor plan analysis
AI quotation engine
Lead scoring & prioritisation
Sales pipeline visibility
Automated customer communication
Quotation review workflows
Operations dashboard & KPIs

Technology stack

Frontend: ReactTanStackTailwind
Backend: Node.jsEdge FunctionsAI workflows
Data: PostgresRow-level securityVector store
Integrations: AI systemsSales automationQuotation automationDocument processingWorkflow automationBusiness intelligence

Architecture

AI Revenue & Quotation Platform — system architecture

Customer conversations, AI qualification, document analysis and quotation generation share a single operating system — designers and management work from the same source of truth.

Client
Service
Data
External

01 · Client surfaces

Customer (WhatsApp)

Natural conversation

Sales console

Leads · pipeline

Management dashboard

Revenue · KPIs

02 · AI & application services

Lead qualification AI

Scoring · fit · intent

Requirements extraction

Scope · budget · timeline

Floor plan analysis

Layout · sqft · rooms

AI quotation engine

Pricing · recommendations

Customer comms service

Follow-ups · reminders

03 · Data & integrations

Postgres database

Leads · quotes · history

Pricing rules engine

Margins · constraints

Document store

Floor plans · quotes

Messaging API

WhatsApp · email

Workflow

Enquiry to approved quotation — manual vs AI-automated

What used to take days of manual sales work now runs as a structured AI pipeline.

Manual processHigh friction · human-in-the-loop

01
Enquiry sits in personal inbox
Slow, inconsistent response
02
Designer manually qualifies lead
Quality varies by person and day
03
Floor plans reviewed by hand
Time spent before knowing fit
04
Quotations prepared from scratch
Hours per quote, easy to drift
05
Follow-ups tracked in heads & chats
Revenue lost to missed prospects

Automated systemSingle system · runs on its own

01
Enquiry captured & acknowledged instantly
Customer engaged within seconds
02
AI qualifies lead, fit and intent
Consistent scoring on every enquiry
03
Floor plans analysed automatically
Scope and sqft extracted up front
04
AI generates quotation recommendation
Designer reviews, doesn't author
05
Pipeline prioritised & follow-ups automated
Management sees the full picture

Outcomes

What changed for the business.

Faster lead response times
Customers engaged in seconds, not hours.
Consistent lead qualification
AI scores every enquiry to the same standard.
Reduced administrative workload
Designers spend time on projects, not paperwork.
Faster quotation turnaround
Quotes prepared by AI, reviewed by humans.
Pipeline visibility for management
Live view of lead flow, conversion and projected revenue.
Scalable sales operations
Grow revenue without proportionally growing headcount.

Case 02 · Healthcare · Clinic Software

Healthcare Scheduling & Patient Operations Platform

Custom healthcare operations software built to streamline appointment scheduling, intake workflows, recurring services, approvals, patient management, and administrative operations.

In productionClient-owned codebase

Challenge

As appointment volume increased, the client relied on fragmented communication channels, manual scheduling processes, and repetitive administrative work. Operational complexity grew faster than the systems supporting it, creating bottlenecks across scheduling, approvals, and patient interactions.

Solution

Custom healthcare operations platform combining appointment scheduling, intake workflows, recurring booking management, approval processes, patient administration, and automated communications within a single system.

Shipped features

Appointment scheduling
Availability management
Multi-service booking workflows
Recurring appointment logic
Intake and onboarding forms
Administrative controls
Automated notifications
Customer portal
Workflow approvals
Operational reporting

Technology stack

Frontend: ReactTanStackTailwind
Backend: Node.jsEdge FunctionsREST
Data: PostgresRow-level security
Integrations: Email notificationsBooking automationApproval workflowsCustomer communicationsAdministrative operations

Architecture

Healthcare Scheduling & Patient Operations Platform — system architecture

Patient-facing app and clinical admin console share a single backend with role-based access, a Postgres source of truth and automated messaging on top.

Client
Service
Data
External

01 · Client surfaces

Patient mobile app

Bookings · doses

Clinical admin console

Records · approvals

02 · Application services

Auth & roles

Patient · staff · admin

Booking service

Availability · approvals

Clinical records

Patient history

Notification engine

Reminders · alerts

03 · Data & integrations

Postgres database

Single source of truth

Audit log

Critical actions

Transactional email

Verification · receipts

Messaging API

Outbound reminders

Workflow

Booking & approvals — manual vs automated

What used to take days of back-and-forth now runs as a structured workflow inside one system.

Manual processHigh friction · human-in-the-loop

01
Patient sends WhatsApp message
Lands in personal inbox
02
Staff types details into spreadsheet
Risk of typos & duplicates
03
Approval requested over chat
Often waits 1–3 days
04
Manual email reminder typed out
Sometimes forgotten
05
No central record of interaction
History lives in chat

Automated systemSingle system · runs on its own

01
Patient books in mobile app
Identity verified by email
02
Request lands in clinical inbox
Structured, validated data
03
Approval workflow with roles
Tracked, time-stamped
04
Automated confirmations & reminders
Email + messaging
05
Full audit trail in patient record
Searchable history

Outcomes

What changed for the business.

One operational system
Replaces WhatsApp, spreadsheets and paper forms.
Approvals as a workflow
Time-stamped, role-based, fully traceable.
Automated patient comms
Reminders and confirmations run on their own.
Full ownership
Full codebase, database, infrastructure, workflows, and deployment environment owned and controlled by the client.
Role-based access
Patients, staff and admins each get the right surface.
Scalable foundation
New clinical modules can plug into the same backend.

Case 03 · Fintech · Financial Services · Education

Financial Modelling & Investment Simulation Platform

In productionClient-owned codebase

Challenge

Most investment simulation and financial education tools are fragmented across spreadsheets, charting platforms and disconnected market data sources. Users struggle to model scenarios, test strategies and understand outcomes in a single system — making structured training, portfolio analysis and decision evaluation slow, inconsistent and difficult to scale.

Solution

Visileo designed and developed a unified investment simulation platform that combines portfolio management, market analytics, scenario modelling and trading workflows into a single environment. Users can build and monitor portfolios, simulate investment scenarios, analyse market behaviour, test trading strategies, track performance over time, explore transaction histories and evaluate investment decisions using structured data — all from one professional-grade workspace.

Shipped features

Portfolio management
Investment simulation
Scenario modelling
Market analytics
Trade tracking
Recommendation engine
Performance reporting
Blockchain transaction explorer
Historical analysis
Risk assessment

Technology stack

Frontend: ReactTanStackTailwindData visualisation
Backend: Node.jsEdge FunctionsREST
Data: PostgresTime-series storeRow-level security
Integrations: Financial market APIsBlockchain explorersPricing feedsAnalytics pipelines

Architecture

Financial Modelling & Investment Simulation Platform — system architecture

Portfolio surfaces, simulation services and market data share a single financial operating environment — analysts, traders and learners work from the same source of truth.

Client
Service
Data
External

01 · Client surfaces

Portfolio dashboard

Positions · P/L

Stock & market research

Charts · fundamentals

Scenario workbench

Modelling · forecasts

Blockchain explorer

On-chain activity

02 · Application services

Simulation engine

Trades · scenarios

Analytics service

Performance · risk

Recommendation engine

Signals · ranking

Reporting service

Statements · exports

03 · Data & integrations

Postgres database

Portfolios · trades

Market history store

Prices · indicators

Financial market APIs

Equities · indices

Blockchain data APIs

Transactions · wallets

Workflow

Investment analysis — spreadsheets vs unified platform

Replaces fragmented spreadsheet and charting workflows with a structured investment environment.

Manual processHigh friction · human-in-the-loop

01
Portfolios tracked in spreadsheets
Manual entry, version drift
02
Market data pulled from multiple sources
Inconsistent, hard to reconcile
03
Scenarios modelled ad-hoc
Not repeatable or comparable
04
Strategy testing done by hand
Slow, error-prone, unscalable
05
Performance reviewed after the fact
No live visibility or audit trail

Automated systemSingle system · runs on its own

01
Portfolios managed in one platform
Live positions, P/L and exposure
02
Market data unified into the environment
Consistent, queryable, historical
03
Scenarios modelled with structured inputs
Repeatable, comparable, shareable
04
Strategies simulated against real conditions
Test before committing capital
05
Performance tracked continuously
Recommendations and risk surfaced in real time

Outcomes

What changed for the business.

Replaced spreadsheet-driven analysis
Investment workflows moved into a structured platform.
Centralised the financial toolset
Portfolios, analytics, scenarios and trading in one system.
Improved portfolio visibility
Live P/L, allocations and exposure at a glance.
Enabled structured investment training
Repeatable scenario and strategy testing for learners and analysts.
Faster, more confident decisions
Evaluate strategies against real market behaviour before committing capital.
Scalable financial foundation
A platform built to extend into future financial products.

Case 04 · Luxury Retail · Commerce & Operations

Luxury Commerce & Operations Platform

Custom commerce operations platform built for a luxury direct-to-consumer retail brand.

In productionClient-owned codebase

Challenge

The business relied on fragmented manual processes across orders, customer communications, inventory management, payment verification, and fulfilment operations. As order volume increased, administrative overhead grew and visibility across the business became increasingly difficult.

Solution

Designed and implemented a unified commerce operations platform that centralised customer management, order processing, inventory oversight, payment workflows, operational reporting, and marketing activities into a single system.

Shipped features

Order management
Payment verification workflows
Customer management
Product and inventory tracking
Revenue analytics
Abandoned cart recovery
Marketing and campaign operations
Administrative back-office workflows
Executive dashboard & KPIs
Stock alerts & thresholds

Technology stack

Frontend: ReactTanStackTailwind
Backend: Node.jsEdge FunctionsREST
Data: PostgresRow-level security
Integrations: Commerce platformOperations managementRevenue analyticsCustomer managementWorkflow automation

Architecture

Luxury Commerce & Operations Platform — system architecture

Admin console and storefront share a unified commerce backend — orders, inventory, payments, customer records and marketing operations in one system.

Client
Service
Data
External

01 · Client surfaces

Luxury storefront

Catalog · checkout

Customer accounts

Orders · history

Admin console

Operations · analytics

02 · Application services

Catalog service

Products · variants

Order pipeline

Cart → fulfilment

Customer CRM

Profiles · history

Marketing & recovery

Campaigns · carts

03 · Data & integrations

Postgres database

Commerce · CRM

Inventory ledger

Stock · movements

Payment workflow

Verification · logging

Transactional email

Receipts · campaigns

Workflow

Commerce operations — manual vs automated

Order, inventory, payment and customer data stay in sync from checkout to fulfilment.

Manual processHigh friction · human-in-the-loop

01
Orders processed across messages and spreadsheets
No central record, easy to miss
02
Payment verification handled manually
Screenshots, delays, reconciliation issues
03
Inventory checked in separate documents
Often out of date before it is viewed
04
Customer data retyped for every order
No persistent history or profiles
05
Marketing campaigns tracked outside the system
No visibility into performance

Automated systemSingle system · runs on its own

01
Orders flow through the unified platform
Structured, tracked, searchable
02
Payment verification integrated into workflow
Logged and confirmed automatically
03
Inventory updates in real time with every order
Stock alerts when thresholds hit
04
Customer profiles build automatically
Order history, preferences, activity
05
Marketing and recovery campaigns run from one place
Abandoned carts, signups, promotions

Outcomes

What changed for the business.

Centralised business operations
Orders, customers, inventory, payments and marketing in one system.
Reduced manual workload
Automated verification, tracking and reporting replaces repetitive admin.
Faster order processing
Structured workflows from checkout to fulfilment.
Improved revenue visibility
Live analytics on revenue, conversion and customer activity.
Better inventory oversight
Real-time stock levels with automated threshold alerts.
Scalable foundation
Operational infrastructure built to support growth.

How We Replace Manual Ops

From scattered process to single system.

The shape every Visileo build takes — regardless of industry.

Workflow

Manual operations → automated platform

The same pattern repeats across healthcare, retail, fintech and HR work.

Manual processHigh friction · human-in-the-loop

01
Work coordinated in chat
WhatsApp · email · DMs
02
Data lives in spreadsheets
Versions, tabs, drift
03
Approvals over messages
Untracked, slow
04
Manual data re-entry
Across 2–4 tools
05
Reporting after the fact
Stale by the time it lands

Automated systemSingle system · runs on its own

01
Structured workflows in-product
One surface for the team
02
Single source of truth
Postgres-backed
03
Role-based approval pipelines
Time-stamped, auditable
04
Integrated automations
Email · messaging · webhooks
05
Live operational dashboards
Decide on current data

Reference Architecture

The shape of a Visileo build.

Most platforms we ship follow this layered architecture — adapted to the domain, owned end-to-end by the client.

Architecture

Production reference architecture

Client surfaces, application services, data and integrations — typed contracts between every layer.

Client
Service
AI
Data
External

01 · Client surfaces

Web app

React · TanStack

Mobile app

Native-feeling

Admin console

Operations UI

02 · Application services

Auth & roles

RBAC · sessions

Domain API

Typed contracts

Background jobs

Queues · cron

AI workflows

Agents · automations

03 · Data & integrations

Postgres

Source of truth

Audit log

Traceability

Email · messaging

Transactional

Payments

Where required

More Projects

A wider sample of the work.

Each is a custom platform with its own architecture, stack and operating model.

05 · HR Tech · Workforce

Workforce Management System

Workforce platform unifying employee records, scheduling, timesheets and operational reporting.

System features

Employee records
Scheduling
Timesheets
Admin reporting

Stack

React
Node.js
Postgres
Role-based auth

Architecture

Records · scheduling · reporting modules on a shared admin backend.

06 · SaaS · Operations

Membership Operations Platform

Member management, class bookings, billing and a multi-location operations dashboard.

System features

Member management
Class bookings
Billing workflows
Multi-location

Stack

React
Node.js
Postgres
Edge Functions

Architecture

Member + bookings + billing services on a single multi-tenant backend.

Capabilities

Across the stack.

Frontend

React
TanStack
Tailwind
React Native
Vite

Backend

Node.js
Edge Functions
Postgres
REST · GraphQL
Webhooks

AI & Infra

OpenAI · Anthropic
Vector DBs
Cloud Workers
Background jobs
Realtime

Start a project How we build

Engagement

How We Engage

Every project starts with a discovery call. We scope the right tier for your stage — no fixed packages, no shelf-ware.

Starter Systems

From $2k+

Built from proven foundations, customised, branded and configured for your business.

SMEs and early-stage businesses

Template-based or white-label foundation
Custom branding, content and configuration
Essential integrations and automations
Launch-ready within weeks

Start with a Starter System

Sprint

$10k – $25k

Focused software initiatives that solve a specific operational problem fast.

Internal tools, portals, workflow automation

Scoped to 2–4 week delivery
Single-module or proof-of-concept
Real users onboarded early
Clear path to expand into a full build

Plan a Sprint

Build

$25k – $75k

Custom platforms designed around your workflow with multi-role access and integrations.

Custom platforms, multi-role systems, integrations and automation

End-to-end design and engineering
Multi-role auth and approval pipelines
Third-party integrations (payments, messaging, ERP)
Production launch with observability

Scope a Build

Platform

$75k – $250k+

Core business systems that operate as long-term competitive infrastructure.

Complex workflows, multiple integrations, long-term product development

Architected for scale and longevity
Multiple integrated services and surfaces
AI and automation as first-class citizens
Dedicated roadmap and iteration cycles

Discuss a Platform

Next step

Not sure which tier fits? Most clients don’t on day one. We’ll map your operations, identify the bottleneck and propose the right engagement shape — usually within one call.

Schedule a Discovery Call

FAQ

Common questions. Straight answers.

Everything you need to know about working with Visileo — from pricing and timelines to ownership, integrations and government support.

Most projects start between SGD 15,000 and SGD 80,000 depending on scope, complexity and integrations. We scope every engagement transparently after a discovery session — no hidden fees, no scope creep without discussion.

Yes. Every client receives full ownership of the codebase, documentation and deployed infrastructure. We build on your accounts where possible, and hand over everything on completion. No licensing fees, no lock-in.

Always. We begin with an operational audit to map workflows, identify bottlenecks and validate the business case. You receive a clear roadmap and fixed-scope proposal before any build begins.

We design around your current stack. Our platforms integrate with existing CRMs, accounting tools, payment providers, WhatsApp Business APIs and any system with an accessible interface.

Yes — from AI-powered lead qualification and automated quotation engines to workflow automation and intelligent document processing. We identify where AI creates genuine operational leverage, not novelty.

Healthcare, fintech, luxury retail, professional services, workforce management and SaaS. We bring sector-specific reference architectures and proven patterns to every engagement.

Starter systems ship in 4–6 weeks. Full custom platforms typically take 8–16 weeks. We work in fixed-milestone sprints with weekly progress reviews, so timelines are predictable.

Yes. We help Singapore-based clients evaluate and apply for relevant government grants and support schemes as part of our engagement process, reducing net project cost.

Start a project

Ready to build something exceptional?

Whether you need a lean Starter System or a full custom platform, we'll help scope the right build for your stage.

Schedule a Discovery Call Explore Work