TELIC

Our Work

See how we think and build.

These demonstrations show the types of operational systems we design and build. They use synthetic data and are not presented as client case studies. Each one is labelled for exactly what it is, and client work will be published here only where clients have authorised it.

Labels used:ConceptPrototypeInteractive DemoDeployed ProductClient Project

Illustrative software examples. Dashboard, order and stock data are synthetic.

Project 01Prototype

Smart Business Operations Platform

A business management system that lets a growing SME run its core operational processes — inventory, products, suppliers, customers, orders and purchasing — from one application with a consistent source of information.

Business challenge

Growing businesses often struggle to manage operations across spreadsheets, paperwork and disconnected software. Stock counts, orders and supplier details drift apart, and nobody has a single reliable view.

Solution

A unified operational platform that centralises the important processes, applies business rules consistently, controls who can see and change what, and reports on the whole operation.

Features

Inventory & stock movement trackingProduct & supplier managementCustomer & order managementPurchasingUser roles & permissionsOperational dashboards & reportingActivity logs

What it demonstrates

Full-stack engineeringSystem architectureRelational database designAuthentication & authorisationBusiness rulesData validationReporting

Intended users

Operations managers, warehouse and sales staff at a 10–100 person business; the owner for reporting.

Intentionally not included

Accounting (it integrates with yours rather than replacing it), payroll, multi-company consolidation.

A real implementation would need

Discovery of your actual order and stock flows, data migration from current spreadsheets, user training and a phased cut-over.

Technology

ReactTypeScriptNode.js & ExpressPostgreSQLREST API
Operations Platform — interactive preview
Open orders4712 awaiting stock
Stock valueR 1,28 m318 SKUs
Below reorder93 on purchase order
Orders this week
Mon
Tue
Wed
Thu
Fri
09:42SO-1042 packed by Warehouse (T. Mokoena)
09:15PO-311 raised for Strapping seals 16mm
08:58Stock count adjusted: Thermal labels −4

Select a section to explore. Interface preview with synthetic demonstration data.

Simplified architecture

Web clientAPI & business rulesRelational database
Auth, roles & audit log
Results: internal demonstration project. Implemented capabilities and verified testing results are published here as they are completed — no client outcomes are claimed.
Project 02Concept

Intelligent Document Processing System

An AI-powered application that reads business documents — supplier invoices, delivery notes, purchase orders, receipts — and extracts structured information with confidence indicators and human validation before anything is passed on.

Business challenge

Businesses spend significant time manually capturing and verifying document information, and errors in that capture flow downstream into stock, payments and reporting.

Solution

An intelligent document-processing workflow that reduces repetitive capture while keeping a person in control: the system proposes, a human approves, and only approved data moves to other systems.

Features

Document upload (PDF & image)OCR processingAI-assisted field extractionConfidence indicatorsManual validation stepProcessing status & historyApproved data transfer to other systems

What it demonstrates

Python backend engineeringAI integrationOCRData processingREST APIsValidation workflowsHuman-in-the-loop AI design

Intended users

Accounts-payable and admin staff who capture supplier documents; a manager who approves exceptions.

Intentionally not included

Fully automatic posting with no review step — by design.

A real implementation would need

A sample of your real documents to measure variation, agreed validation rules, and a connection to your accounting or stock system.

Technology

PythonOpenCVTesseract OCRLLM extraction APIPostgreSQL
Document Processing — Review
Coastal Packaging SuppliesTAX INVOICE
INV-20418 · 14 Oct 2026
VAT no. 41 082 85763
Pallet wrap 500mm24R 2 976,00
Thermal labels 100×15040R 1 480,00
Strapping seals 16mm10R 645,00
Total incl. VATR 5 866,35
Supplier98%Coastal Packaging Supplies
Invoice number99%INV-20418
Invoice date97%14 Oct 2026
VAT number62%41 0_2 85763Low confidence — please confirm
Total incl. VAT96%R 5 866,35
ApproveEdit fields

Simplified architecture

UploadOCRAI extractionHuman reviewApproved data → systems
Results: concept stage. Capabilities and testing results will be published as they are implemented and verified.
Project 03Concept

Connected Business Automation Hub

A system that shows how business applications can exchange information and trigger automated workflows — with the retries, logging and monitoring that make integrations trustworthy in day-to-day use.

Business challenge

Applications that can't exchange information force staff to re-key data and chase status updates by email, and integrations built without monitoring fail silently.

Solution

An event-based hub that receives business events, runs configured workflows, synchronises data between systems and reports clearly on what ran, what failed and what was retried.

Example workflow

  1. 1A customer order is created.
  2. 2The inventory system receives the order information.
  3. 3Stock availability is checked.
  4. 4An approved workflow updates the order status.
  5. 5Relevant staff receive notifications.
  6. 6Authorised financial information is transferred to a connected system.

Intended users

Whoever currently chases status by email; IT or an owner who needs to see what ran and what failed.

Intentionally not included

Replacing any of the connected systems; this works with what you have.

A real implementation would need

An inventory of your systems and their integration options, agreed ownership of each data item, and monitoring someone actually reads.

What it demonstrates

API architectureAsynchronous processingSystems integrationWorkflow automationLogging & monitoringError handlingData consistencyIntegration security
Automation Hub — Execution log
128 runs today126 succeeded2 retrying
order.createdShop → InventorySucceeded0.8 s
stock.checkedInventory → OrdersSucceeded0.4 s
invoice.approvedDocs → AccountingRetry 1 of 3—
order.status_changedOrders → NotificationsSucceeded0.3 s
supplier.updatedCRM → InventorySucceeded1.1 s
Concept mockup with synthetic data.

Simplified architecture

Source systemsEvent queue & workflow engineTarget systems
Logs, retries & monitoring
Results: concept stage. Capabilities and testing results will be published as they are implemented and verified.

How the projects connect

More than isolated applications.

The three systems are designed to work together. A supplier invoice is processed by the AI document system, reviewed by an authorised user, then passed through an integration workflow into the operations platform — demonstrating connected business systems, not just separate apps.

Step 1AI Document ProcessingReads the invoice, proposes fields, flags uncertainty. A person reviews and approves.
Step 2Automation HubReceives the approved data as an event and runs the integration workflow with logging and retries.
Step 3Operations PlatformRecords the purchase, updates stock and supplier records, and reflects it in reporting.

All demonstrations use synthetic data. Prototype functionality is clearly distinguished from completed features.

Want to see how this would apply to your business?

We'd rather talk about your processes than our portfolio. Tell us what's slowing you down.

Discuss Your Challenge

Telic · Information

More information