Travel agency

An AI trip planner that turns website visitors into quotations

  • Retrieval-grounded assistant
  • Lead capture
  • Human approval
The problem
Enquiries arrived at all hours, staff spent their day answering the same questions, and drafting each itinerary by hand slowed every sale.
What we built
A website assistant grounded in the agency’s own destinations and packages. It understands what the traveller wants, verifies their email and phone, and drafts a day-by-day itinerary. The sales team prices it in a dashboard, the system emails a PDF quotation, and follow-ups are scheduled automatically.
What changed
Every enquiry is answered instantly and reaches the sales team as a qualified lead with a draft itinerary attached. Prices are never shown until a person approves them.
Training institute

A career-guidance assistant that recommends the right course

  • Guided assistant
  • Course recommendations
  • Low-cost hosting
The problem
Prospective students wanted to know whether a course would improve their earnings. Counsellors could not answer every visitor, and good leads went cold.
What we built
An assistant on the institute’s existing website that asks about a visitor’s skills and experience, shows expected salary ranges across city tiers, highlights skill gaps, and recommends courses from the institute’s own catalogue before handing verified leads to a counsellor.
What changed
Visitors get useful, personal guidance on the spot, and counsellors receive warm, verified leads. It runs on ordinary shared hosting with low monthly running costs.
Manufacturing

An offline-first ERP for planning, inventory and shop-floor tracking

  • On-premise
  • Inventory and planning
  • AI-ready data
The problem
Project planning, stock and production status lived in spreadsheets and phone calls. The factory could not depend on a reliable internet connection.
What we built
An on-premise system for project planning, inventory and process tracking that works fully offline, with QR codes to track work on the shop floor and a clean data foundation ready for AI forecasting later.
What changed
One source of truth for what is planned, in stock and in progress, running on the factory’s own hardware.
In-house platform

A private AI assistant that never sends data outside

  • Private AI
  • Agentic retrieval
  • On-premise
The problem
Many organisations want to ask questions of their documents and databases, but cannot send confidential data to a public AI service.
What we built
A fully local, agentic assistant that searches documents, queries databases and remembers context, running on a local language model. It can be used from a web app, an embeddable website widget or Telegram.
What changed
The foundation we use when a client needs AI on sensitive data: everything stays on their own machines.
Healthcare education, Europe

A GDPR-first mobile learning platform for nursing trainees

  • Mobile app
  • GDPR by design
  • International client
The problem
A German provider needed a mobile learning and community app for trainee nurses, under strict European data protection rules.
What we built
An iOS and Android app with an admin panel, designed around GDPR from the first line: EU-only data storage, no personal data in logs, and careful handling of deletion and consent.
What changed
A platform that meets European compliance expectations, built and supported from India.
Public sector

Applied AI training for government officials

  • Training
  • Safe AI use
  • Government
The problem
Officials were hearing about AI everywhere but had no practical sense of how to use it safely in everyday administrative work.
What we built
Hands-on training for Class 2 and Class 3 officials using their own kinds of tasks: drafting, summarising, translating and checking AI output, with clear rules on what should never be pasted into a public tool.
What changed
Part of more than 200 AI workflow trainings delivered to teams, institutes and departments.

Ready to put AI to work in your business?

Tell us what is slowing your business down. We will tell you honestly whether AI can help, and what it would take.