Discover TechTahi

TechTahi

Empowering Communities Through AI & Innovation

Minh Nguyen Interactive 2-hour workshop AUT-developed platform

Discover TechTahi

What this 2-hour workshop is about

This workshop introduces TechTahi as a practical platform for turning ideas into working digital tools. The session combines explanation, demonstration, guided practice, and discussion so participants can experience the workflow directly.

What the session includes

A short introduction to TechTahi and its purpose.
A demonstration of prompting and code generation.
Guided hands-on activities using the platform.
Discussion of real-world uses in teaching and creative work.

What participants will experience

How prompts can become working webpages and tools.
How generated HTML can be pasted back into TechTahi.
How results can be previewed, improved, and reused.
How simple ideas can become useful digital outcomes.

What we will build and explore

Letters, calculators, games, charts, image tools, and QR examples.
Prompt enhancement, preview, and local reuse.
Safe local workflows and practical case studies.
Ideas participants can adapt to their own situations.

After this workshop, you can go home with your own ideas and build useful software for your work, study, community project, or business. No coding background, no subscription, and no extra charge: just a ready-to-use workflow you can keep applying to real problems that matter to you.

Why we created TechTahi

Rationale: Why We Created TechTahi

Technology should not be a locked door for a few. It should be a wharenui where everyone can enter and create. TechTahi was created to lower barriers, turn plain-language ideas into working digital solutions, and make digital creation safer, more open, and more culturally grounded.

Open to all

For too long, building software has required climbing a steep staircase of coding, syntax, and jargon. Many people were kept out not because they lacked ideas, but because the tools were not made for them.

Built for safe experimentation

TechTahi changes that. It helps people turn plain-language ideas into working digital solutions, supporting mana motuhake in the digital space and making it safer to experiment, learn, and create without fear.

Low-cost digital wharenui

TechTahi is designed as a low-cost digital wharenui: open to all, built for collaboration, and grounded in culture so innovation reflects who we are, not only what technology can do.

What this means in practice

1

Open to all Transparent, no sign-in, no subscriptions, and fewer barriers for people who want to create.

2

Built for collaboration People can share ideas, refine them together, and celebrate useful outcomes.

3

Grounded in culture Innovation should reflect identity, community, and context, not just technical possibility.

Core idea

Useful software should start with a clear need, not wasted AI cycles

TechTahi helps people turn a well-described idea into a working result, then refine and reuse it without relying on endless regeneration.

Current pitfall to avoid

Token-burning AI agents can keep looping, regenerating, and charging cost before the idea is clear.
Users may pay for repeated attempts instead of owning a simple working result.
Business fit can become secondary to long automated implementation runs.

TechTahi is deliberately different: ask clearly, generate once, preview locally, refine the result, then reuse or share it without turning every change into another expensive AI cycle.

Open to all Built for collaboration Grounded in culture

In the past, software implementation was often the bottleneck in a project. With AI, the bottleneck is shifting toward idea generation, clearly describing what is needed, and making sure the result fits the real business need. We have to recognise this change and learn how to work with it.

Guided workshop task

First activity: get your first working code

For the first guided task, we will ask TechTahi and Gemini to build a live world clock with five popular cities, plus Auckland, Hanoi, and Jakarta.

Step-by-step guide

Go to the main TechTahi page.
In Prompt Builder, enter the world clock request shown on this slide.
Choose Copy Main Query.
Click Gemini first and paste the copied prompt there.
Wait for Gemini to return one HTML code snippet.
Copy the returned code, click Paste Code, then click Preview.
Use Save to store the HTML file on your local drive. At any later stage, you can drag and drop that saved HTML file back into the code output box in TechTahi to continue editing or changing it.

Use this first prompt

Prompt Builder text

Build a live world clock with 5 different popular cities, and also add Auckland, Hanoi, Jakarta

Sample output Open TechTahi

Goal: get the first working HTML code back from the LLM, paste it into TechTahi, and preview it successfully.

If the LLM does not return one single code snippet, send: QUICK answer in a single HTML code block.

What to check after Preview

The page opens successfully in Preview.
Multiple cities are visible on the clock interface.
Auckland, Hanoi, and Jakarta are included.
The clock updates live rather than showing static text.

If the first result is not quite right

Ask Gemini to improve layout, readability, or city labels.
Use Copy Fix-Code Query if the returned code has errors.
Use Edit GUI later if you want to adjust the visual design manually.

Simple and repeatable process

Why the TechTahi workflow is practical

Up to 8 LLMs Users can choose from multiple external models, and most offer a free tier.

Seconds, not hours Paste generated HTML back into TechTahi and run it almost immediately.

Save or share Refine with GUI or code, then keep it locally or share it online.

1. Choose from up to 8 LLMs

TechTahi can work with up to eight available LLM options, and most of them provide a free tier, so many users do not need to pay to get started.

2. Use TechTahi as the free working interface

TechTahi provides the interface and supporting libraries to hold, run, preview, and manage code generated by external LLMs.

3. Reduce cost and setup overhead

There is no need for a complicated installation path, paid hosting setup, or a separate development stack before trying an idea.

4. Ready in seconds, not minutes or hours

Generated code can be pasted back and run almost immediately, so users can move from idea to testable output very quickly.

5. Support semi-coding development

Users can refine outputs through both GUI-level editing and direct code injection using Edit GUI and manual code updates.

6. Save locally or share online

Once the result works, users can save locally for themselves or share online so other people can access and use it.

Main practical message

TechTahi gives users a free front-end workspace around external LLMs.
Most users can begin with free-tier AI tools rather than paid subscriptions.
The platform removes much of the cost, installation, and deployment friction.
It supports both visual refinement and manual code-level adjustment.
Outputs can be previewed, saved locally, or shared with other people online.

The key idea is simple: let external LLMs generate the code, while TechTahi provides the practical environment to hold it, run it, refine it, and share it quickly.

Access principle: TechTahi is designed with the assumption that some participants may have limited or no budget. TechTahi does not ask users for payment, subscription, or paid setup before they can start building useful software.

Practice using Edit GUI

Second activity: improve a home loan calculator with Edit GUI

This activity focuses on semi-coding development. Participants start with a simple home loan calculator, then use Edit GUI to refine the wording, layout, labels, and overall presentation visually.

Suggested practice flow

Ask the LLM to build a home loan calculator webpage.
Paste the generated HTML into TechTahi and click Preview.
Open Edit GUI to adjust headings, colours, spacing, labels, and result areas.
Edit the title and helper text so the task clearly explains W, R, and T in plain language.
Rewrite button text and result wording so users can understand the calculator without technical background.
Apply the changes and preview the improved result again.

Example prompt to start with

Prompt Builder text

Build a clean home loan calculator webpage with input fields for W, R, and T, where W is the affordable weekly payment amount, R is the annual interest rate in percentage, and T is the loan term in years. Include clear labels and a button to calculate the result.

Sample output Open TechTahi

Main learning goal: see how TechTahi supports visual improvement after the first code is already running.

What participants should edit

Rewrite the title and instructions so the home loan task is easier to follow.
Improve labels for W, R, and T so users understand each input clearly.
Change colours, spacing, or button text to improve clarity.
Rearrange sections so the calculator and result area read more clearly.

Why this matters

Shows that users do not need to rely only on prompt regeneration.
Introduces a more practical semi-coding workflow.
Builds confidence in refining real interfaces visually.
Helps users move from generated output to polished result.

Why direct LLM-only use has downsides

Risks and trade-offs of LLM use

Many people use LLMs by chatting to them repeatedly about every problem and asking them to solve each step. Each of those requests requires inference, and inference is not free. It consumes money, computing power, electricity, cooling, water, and time.

TechTahi aims to support an ask once, reuse many times workflow, so users can reduce repeated LLM calls, lower hidden computing cost, and make more responsible use of energy and resources.

What often happens in normal LLM use

Users repeatedly ask the model to explain, revise, fix, and regenerate.
Every new prompt sends another request through the model.
This repeated inference increases cost and infrastructure demand.
It can also encourage dependence on constant back-and-forth prompting.

Why that matters

Large models require expensive hardware and specialised infrastructure.
Cooling those systems can require significant energy and water.
More requests can mean more cost, more waiting, and more environmental load.
Outputs can still be biased, incorrect, opaque, or hard to verify.

What is deep learning?

Use this page to understand the training foundations behind modern AI systems and why the infrastructure behind them is so large.

Open deep-learning.html

What is an LLM and how does it run?

Use this page to explain what happens during generation and why each extra prompt still consumes hidden power and infrastructure.

Open large-language-model.html

What are the costs and risks?

Use this page to discuss the wider cost, risk, and environmental impact behind heavy repeated LLM use.

Open llm-risk.html

These three pages help explain why repeated LLM use can carry hidden power costs, and why TechTahi tries to reduce unnecessary repeat prompting after the first useful result is generated.

Use the good, reduce the bad

Take the good, prevent the bad

TechTahi uses LLMs for fast setup and first-draft generation, but the working rule is simple: sensitive data is not sent to the model. We keep it local and secured, then use the generated tool to handle the real information afterwards.

Harness the strengths of LLMs while reducing risks

Use LLM capabilities for rapid development.
Only query the LLM during the creation and setup phase.
No constant back-and-forth, which reduces exposure and dependency.
Maintain full control, privacy, and independence in deployment.

No sensitive data leakage

User private or proprietary information is never sent to the LLM.
All confidential content stays fully local.
This includes student names, grades, and other restricted records.
The safer pattern is template first, local entry later.

Local execution after setup

Prompt the template first. After the tool is generated, it can run locally on the user device, and in some cases entirely offline, with no need to keep sending information back to an external model.

Safe pattern: create once -> keep sensitive data local -> reuse locally afterwards

Best of both worlds

TechTahi keeps the speed of LLM-assisted creation while reducing privacy risk, repeated interference, and deployment dependence. The result letter example shows this clearly in practice.

Use AI for setup Keep sensitive data local and secured Reuse the tool without repeated exposure

Prompt safely, fill locally

Third activity: build a result announcement letter without sending sensitive data

This activity reinforces the privacy rule. Use the LLM to create the letter tool and layout, but keep the real student data local. The prompt should describe the structure only, and the actual student name and grade should be entered later in the generated page.

Open letter example

Suggested practice flow

Explain that private student data should not be pasted into the LLM.
Ask the LLM to build a result announcement letter page with fields for later local entry.
Paste the generated HTML into TechTahi and click Preview.
Open the result letter tool and type the real student name and grade locally.
Show that different grades can produce different letters without exposing the sensitive data in the prompt.
Emphasise that the completed tool can then be reused locally with more control and less exposure.

Example prompt to start with

Prompt Builder text

Build a result announcement letter for teacher named David Mark, allow to enter student name, e.g. Will Smith, student overall grade, e.g. 75, and a general letter telling him their grade, make sure different grades have different letters. Keep the real student data for local entry after the page is generated.

Sample output Open TechTahi

Main learning goal: use the LLM for rapid setup, but protect sensitive student data by keeping it local and entering it only after the tool is running.

What participants should notice

The LLM helps create the form, structure, and letter logic quickly.
The private values can stay on the device instead of going into the prompt.
The same local tool can be reused for many students after setup.
This reduces exposure, repeated prompting, and unnecessary dependence.

Why this matters

It shows a realistic privacy-conscious workflow for education and administration.
It demonstrates how TechTahi separates generation from local use.
It helps people understand that AI assistance does not require sending confidential content.
It supports restricted, high-security, and offline-friendly use after setup.

Secure local-first workflows

New activity: local storage, encryption, and private file access

This slide highlights a strong TechTahi use case: generating tools that work locally, keep data on the device, and avoid unnecessary internet upload. That makes it useful for private notes, local search, and security-conscious tasks.

Activity 1: secured local diary

Ask the LLM to build a diary that unlocks with a PIN popup.
Require the PIN itself to not be stored directly anywhere.
Store the diary text locally, but encrypt it using the PIN.
Use auto-save so the writing is preserved without sending it online.

Prompt Builder text

Build a secure diary app, on first use, ask the user to create a PIN, hash it, store only the hash, encrypt all diary text using a key derived from the PIN, auto-save the encrypted diary to localStorage, show a PIN popup to unlock, include a lock/logout button.

Sample output Open TechTahi

Activity 2: local text file search system

Build a search tool that works with local drives and folders.
Let users search across text files for words, phrases, or full sentences.
Keep the file reading and searching on the local machine where possible.
Show how TechTahi can support useful desktop-style workflows without cloud upload.

Prompt Builder text

build a comprehensive search system that allow to search local drive and folder for all text file and search for words or sentences in these text files

Sample output Open TechTahi

Why these examples are strong

They show local storage can be more useful when paired with encryption and careful unlock logic.
They demonstrate that sensitive notes can stay on the user device rather than being uploaded to an external service.
They highlight TechTahi as a path to build privacy-aware tools quickly, then run them locally.
They help users understand the difference between generating the tool once and using the tool safely many times.

Security and practical value

Encrypted local diary data gives better protection than plain local storage alone.
Local file search can enhance document access and review without uploading text to the internet.
These activities support privacy-conscious teaching, admin work, and personal productivity workflows.
They reinforce that TechTahi can help create secure and capable local tools, not only public web demos.

Main learning goal: show how TechTahi can help generate tools that keep data local, add stronger protection with encryption, and improve access to files without relying on internet upload.

Use prompt enhancement

Next activity: improve ideas with Copy Enhance Query

This activity shows how TechTahi can strengthen a simple idea before the main code generation step. Start with a short prompt, use Copy Enhance Query, let an LLM expand it, paste that enhanced result back into Prompt Builder, then use Copy Main Query to generate the final code. This also helps when you want to build more complicated systems, not only small demos.

Open arcade sample Open Paint Studio

Suggested practice flow

Put this short prompt into Prompt Builder: Breakout (1976) arcade game with modern look.
Click Copy Enhance Query.
Choose an LLM and let it run the enhance prompt.
Copy the enhanced output back into the Prompt Builder.
Click Copy Main Query and send that to the LLM.
Copy the returned code back into TechTahi and preview the result.
Repeat the same workflow for a more advanced idea like a Microsoft Paint-style system.

Starter prompts

Prompt Builder text

Breakout (1976) arcade game with modern look

Sample output Open TechTahi

More complex system idea

Build a duplication of Microsoft Paint

Sample output Open TechTahi

Main learning goal: see how Copy Enhance Query can turn a short idea into a much stronger final generation workflow, and also help shape more complicated systems such as a Paint-style app.

What participants should notice

A short prompt can be improved before code generation.
The enhance step can add style, structure, and clearer requirements.
The main query becomes stronger after the enhancement step.
The final output is usually more polished than from the short prompt alone.
The same method can be used to plan larger tools with more features and interface controls.

Why this matters

It teaches a practical two-step prompting workflow.
It shows how TechTahi supports better prompting without making users write long prompts from scratch.
It builds confidence in improving results before the final code generation step.
It shows that Enhance Query can also help users move toward more complicated systems such as a Paint-style application.

Draw graphs

Next activities: create charts and live graph outputs

At work, you may need to plot something from data so people can understand the pattern quickly. A picture can be worth a thousand words: for example, turning x, y, z measurements into a surface map, or showing system activity as a live line chart. Here we show what you can do with TechTahi when data needs to become a clear visual story.

Practical prompt examples: "Create a line chart from monthly sales data. Include sample values for 12 months, allow me to edit the data, and show the trend clearly with labels." / "Build a bar chart to compare student marks across five subjects. Include editable sample data, show the average mark, and highlight the highest subject."

Activity 1: 3D mountain surface

Ask for a 3D surface showing a mountain-like elevation map.
Pre-enter 10 sample data points so users can see something immediately.
Let the user replace the sample x, y, z values later with their own data.

Prompt Builder text

Plot a 3D surface showing a mountain-like elevation map based on x, y, z coordinates given by me. Pre-enter 10 sample data points so the chart displays immediately, and allow the sample values to be edited later.

Sample output Open TechTahi

Activity 2: real-time CPU chart

Create a live line chart that updates every second.
Display CPU usage changing over time in a clear visual way.
Use simulated data if real system data is not available.

Prompt Builder text

Create a real-time line chart that updates every second to display CPU usage. Show a clean dashboard style and simulate changing values if real CPU data is not available.

Sample output Open TechTahi

Main learning goal: see how TechTahi can generate both static visualisations and live-updating graph interfaces from plain-language prompts.

Basic image processing

Next activities: upload an image and transform it

At work, you may need to adjust or inspect an image quickly without opening specialist software. A simple visual tool can save time: for example, checking whether a scanned document is clearer in grayscale, applying filters to highlight details, or rotating and flipping an uploaded image for a report. Here we show what you can do with TechTahi when an image needs to become easier to view, compare, or present.

Practical prompt examples: "Create an image cleanup tool where I can upload a scanned document, adjust brightness and contrast, and preview the clearer result." / "Build an image comparison page where I can upload two photos, zoom in, and compare them side by side for small visual differences."

Activity 1: image filters

Allow users to upload an image.
Add grayscale, blue tint, sharpen, and edge detection.
Show the original and processed versions clearly.

Prompt Builder text

Allow the user to upload an image and apply some basic image processing options such as grayscale, blue tint, sharpen, and edge detection. Show both the original and processed image clearly.

Sample output Open TechTahi

Activity 2: translate, rotate, and flip

Allow users to upload an image.
Add translation, rotation, horizontal flip, and vertical flip.
Make the controls easy to test and preview.

Prompt Builder text

Allow the user to upload an image and perform basic translation, rotation, and flip operations horizontally and vertically, with a simple preview area and easy controls.

Sample output Open TechTahi

Main learning goal: show that TechTahi can support practical visual tools, not only calculators and text-based pages.

Other cool stuffs

Next activities: visually rich and interesting demos

At work or in a workshop, you may need a quick prototype that feels visual, interactive, or memorable enough to explain an idea. A small demo can make an abstract concept easier to understand: for example, showing a 3D model, placing a point on a map, or creating a QR code for a handout or event. Here we show what you can do with TechTahi when the goal is to make an idea more engaging and easier to share.

Practical prompt examples: "Create an interactive floor-plan demo where users can click rooms to see notes, photos, and status labels." / "Build a simple location map page where I can enter several addresses, show markers, and list each place with a short description."

3D and map ideas

Show a 3D chessboard floating in space, with realistic lighting and shadows.
Show an Auckland map and where you are on the map with a red dot.

Prompt Builder text

Show a 3D chessboard floating in space with realistic lighting and shadows. Also create a separate idea to show an Auckland map and where you are on the map with a red dot.

Sample output Open TechTahi

Quick utility idea

Allow the user to enter text.
Convert that text into a QR code instantly.
Show a clean and easy-to-use interface.

Prompt Builder text

Allow the user to enter text and convert that text into a QR code instantly with a clean interface and a clear preview area.

Sample output Open TechTahi

Main learning goal: inspire participants with a wider creative range, from useful tools to visually impressive prototypes.

Human-centred impact

Educational and cultural value

For teachers, TechTahi outputs can become interactive classroom objects, not just separate web pages. Here I will show how to add these interactive tools into PowerPoint so they can support explanation, demonstration, and student engagement during a presentation.

Educational impact

Makes complex ideas easier to demonstrate interactively.
Supports teaching examples such as Tower of Hanoi and recursion visualisation.
Encourages exploratory learning with immediate feedback.
Useful for classroom engagement and project-based learning.

Cultural impact

Supports culturally grounded and identity-aware digital projects.
Can be extended to Māori motif and storytelling applications.
Helps link technology with language, community, and local relevance.
Encourages more inclusive forms of digital participation.

Activity: Tower of Hanoi

Prompt Builder text

Hanoi Tower animation with input from 1 to 8 rings

Sample output

Activity: Recursive natural trees

Prompt Builder text

Build natural looking trees using recursions

Sample output

PowerPoint workflow: download the TechTahi HTML file, open PowerPoint, then drag and drop the file onto the slide where you want it to appear. To make it clickable during presentation, select the inserted item and use this path:

Insert -> Action -> Object action -> OK

Beyond code - richer creative experience

TechTahi: ready-made images, audio, and icons

TechTahi can support a richer result than code alone. While many LLM workflows mainly return code, TechTahi can now work with ready-to-use images, sound files, and transparent icons so interfaces feel more visual, polished, and interactive from the start.

Not just code, but full media support

Other LLMs typically return only code and not ready-to-use media assets.
TechTahi provides images, audio files, and icons alongside the code workflow.
This helps generated applications feel more engaging and complete from day one.

Dynamic, updatable media libraries

Includes curated images, transparent icons, and audio clips that can be used instantly in generated GUIs.
The libraries can be updated over time as new assets are requested.
New visuals, symbols, and sounds can be added without retraining the model.

Beautiful, ready-made GUI assets

Speeds up interface creation with professional-looking visuals, icons, and sound cues.
Makes generated apps feel more polished, attractive, and interactive.
Supports continuous improvement of the asset libraries over time.

Prompt activity

Prompt Builder text

give me a gallery of many animals with photos, matching icons, soft sound buttons, and short description of them

Sample output Open TechTahi

Main learning goal: show that TechTahi can support media-rich experiences with images, icons, and audio alongside generated code.

Strengths and realistic boundaries

Can TechTahi build complex systems?

Yes and no. TechTahi is excellent for small, functional applications and rapid prototypes. Larger systems can start there, but eventually benefit from modular design and conventional software engineering discipline.

Where it is strongest

Small working apps.
Interactive teaching tools.
Rapid prototypes and proof-of-concepts.
Fast experimentation and sharing.

Where engineering still matters

Complex logic benefits from modularisation.
Databases, multi-user systems, and larger platforms add difficulty.
Long-term maintainability still needs structured design.
TechTahi accelerates early creation but does not replace all software engineering practice.

Prompt activity

Prompt Builder text

Build either a Snake Quiz Game or a Pac-Man Quiz Game to teach primary school students basic arithmetic. The player should collect or eat the correct answer to a random maths problem. Cover addition, subtraction, multiplication, and division. Store the score of the last attempt in local storage, include statistics and levels, and allow the student to progress over time.

Snake sample Pac-Man sample Open TechTahi

Why this example matters

Shows how TechTahi can build small but engaging educational games.
Combines interaction, movement, and arithmetic practice.
Demonstrates that rapid prototypes can still be playful and useful.
Fits well with primary-school teaching and classroom engagement.
Shows that local storage can keep progress, scores, and simple statistics locally without an online database.
Supports storing information locally and securely on the user device.

TechTahi
Empowers
People

Concluding message

TechTahi is more than a code generator. It is a practical access tool for students, teachers, and community users who want to build useful digital resources without needing a coding background or a premium subscription to famous LLMs.

What TechTahi makes possible

Lowers the cost barrier so students, teachers, and non-premium users can experiment and build.
Helps people turn classroom, study, work, or community ideas into working tools.
Makes each step visible: prompt, generate, paste, preview, edit, save, and share.
Keeps users in control of the development workflow instead of hiding decisions inside an agent.
Allows many generated tools to run locally after setup, supporting reuse without constant regeneration.

Why this matters for access

Very good for students and teachers who need practical tools but may not have paid AI accounts.
Useful for anyone who does not have access to premium subscriptions for well-known LLM platforms.
No unnoticed token burning from long-running coding agents quietly trying many hidden steps.
Users decide when to call an LLM, when to reuse code, and when to stop.
The result can be inspected, edited, saved, and shared in a straightforward way.

Final idea: TechTahi makes AI-assisted creation more reachable. Use LLMs to help with creation, but keep the workflow transparent: see the steps, control the cost, avoid hidden token burn, and stay in charge of the result, the data, and local reuse afterwards.

Think about your own situation

More practice: case study ideas and useful prompt rules

The next step is to think about your own work, teaching, or community situation. Choose a real problem, turn it into a small case study, and then write prompts that are clear enough for TechTahi and an LLM to help you build a useful first version.

Case study ideas

A teacher wants a classroom quiz or a subject-specific learning game.
An administrator wants a local form, calculator, or letter generator.
A community group wants a visual information page or local resource tool.
A student wants an interactive project to explain a difficult concept.

Questions to ask yourself

Who is the user?
What problem are they trying to solve?
What inputs are needed?
What should the result look like?
Should any information stay local and private?

Rules for useful prompts

State the goal clearly.
Name the kind of webpage or tool you want.
List the important inputs and outputs.
Say what the user should be able to do.
Mention style or layout only if it matters.
Keep sensitive data out of the prompt.

Simple prompt pattern

Template

Build a [type of tool] for [who it is for]. Include [important inputs]. Show [important outputs]. Make it [style or behaviour]. Keep [sensitive data] local.

Start small, make one useful thing, test it, then improve it.

More prompt ideas to try

One last slide: sample prompts for further practice

TechTahi can generate many impressive single-page apps, not just simple forms. Start with the dashboard example, then try the earlier games, utilities, graph, language, and teaching prompts below.

Dashboard showcase

Student CSV Analytics Dashboard

Upload a student marks spreadsheet and instantly see class trends, risk areas, grade bands, and summary charts. Sample csv file is attached.

Download CSV Sample output Open TechTahi

Tip: you can drag and drop the sample CSV document into the prompt area before generation. This augments the request with real column names and example data, so the generated system can create more relevant filters, charts, summaries, and risk indicators.

Games and interactive tools

Countdown Timer

Build a responsive countdown timer webpage with start, pause, reset, and preset time options. Show a large clear timer display, progress animation, and a clean modern interface for desktop and mobile.

Open TechTahi

Fireworks Sky Animation

Build a full-screen fireworks sky animation with colourful bursts, smooth motion, and a dark night background. Include controls to restart the animation and adjust the number of fireworks.

Open TechTahi

Colorful Bouncing Balls

Create a colourful bouncing balls animation with different ball sizes, speeds, and colours. Make the motion smooth and responsive, and allow the user to add more balls or restart the scene.

Open TechTahi

Sudoku Game

Build a Sudoku game webpage with a clean responsive interface, puzzle validation, mistake checking, and restart options. Include an easy-to-read grid and controls suitable for students and casual players.

Open TechTahi

Minesweeper, Chess, Tic Tac Toe, Tetris

Build a classic Minesweeper game with 3 difficulty levels, or a two-player chess game with move validation, GUI, and history log, or a responsive Tic Tac Toe game, or a playable Tetris game with score, restart button, and clean responsive layout.

Open TechTahi

Horse Racing Simulation

Build a horse racing simulation system with multiple horses, animated race progress, live position updates, lap or distance tracking, and a clear scoreboard. Include start, reset, and race settings controls in a responsive interface.

Open TechTahi Sample output

Audio, speech, and utility prompts

Text to Speech Converter

Build a Text to Speech Converter with a textarea input, voice selection, speaking speed control, and a play button. Make the interface responsive and easy to use for demonstrations.

Open TechTahi

Voice to Text Converter

Build a Voice to Text Converter that records speech from the microphone and shows the recognised text in a clear responsive interface. Include start, stop, and clear buttons.

Open TechTahi

Audio Frequency Player

Build a responsive GUI to input, validate, and play audio frequencies between 20 Hz and 20,000 Hz using Web Audio. Include presets, validation messages, and a clean educational interface.

Open TechTahi

Age Verification Form

Build a responsive age verification form with date of birth input, validation, clear success or failure messages, and a simple professional interface suitable for a public website.

Open TechTahi

Word Cloud Generator

Build a Word Cloud Generator that allows the user to paste text and generate a visual word cloud. Include controls for colours, font size range, and a responsive preview area.

Open TechTahi

Graphs, data, and learning

Tax and Population Graphs

Build a tax rate graph, or display a bar graph of the top 10 ethnic populations in New Zealand, or create a population chart of continents. Use clear labels, colour-coded bars or lines, and a responsive chart layout.

Open TechTahi

Compound Interest and Calculator

Build an investment value calculator using compound interest with input fields, instant calculation, and a graph showing growth over time. Also make it suitable as a simple graphic calculator interface.

Open TechTahi

Scatter Plots and Line Graph

Build a responsive data visualisation page that can show scatter plots and line graphs from user-entered data. Include sample data, editable inputs, axis labels, and a clean chart layout.

Open TechTahi

Culture and language

Learn Maori Quickly

Build a simple webpage to help users learn Maori quickly with useful vocabulary, pronunciation hints, common phrases, and quick revision cards. Make the interface friendly and easy for beginners.

Open TechTahi

Local Learning Helper

Build a local learning helper for vocabulary and revision with flashcards, progress tracking in local storage, and a clear responsive interface for students.

Open TechTahi

Visual Teaching Page

Create a visual teaching page for a classroom topic with short explanations, diagrams, interactive examples, and a layout that works well on both desktop and mobile screens.

Open TechTahi

Tip: start with one short idea, then use Copy Enhance Query to make the prompt stronger before generating the final code.

What TechTahi can help you build

Appendix: practical tools TechTahi can generate

TechTahi can turn a plain-language idea into a working, interactive teaching or productivity tool. It is useful for classroom activities, local data exploration, visual explanation, practice games, reports, and small custom apps that match a specific learning task.

1. Data dashboards

Example: upload a student marks spreadsheet and instantly see class trends, risk areas, grade bands, and summary charts.

Student marks, publication, workload, budget, and survey dashboards.
CSV/Excel upload with chart summaries, heatmaps, scatter plots, and word clouds.

2. Coding and prompt tools

Example: create a prompt builder that helps students structure a task, compare outputs, and save a clean final prompt.

Live editors, formatting helpers, markdown previewers, and safe preview tools.
Prompt builders, code snippet managers, and "explain this code" teaching interfaces.

3. Diagrams and visuals

Example: ask for a research-methodology diagram tool where students enter stages and export a finished visual.

Flowchart, UML, ERD, research methodology, algorithm, timeline, and thesis visualisers.
Diagram export to PNG or PDF.

4. Image and drawing tools

Example: build a poster maker where learners drag labels, colours, notes, and images into a printable layout.

Colouring apps, poster makers, certificate designers, annotation tools, and whiteboards.
Drag-and-drop label activities, mind maps, and motif design tools.

5. Camera, QR, and vision

Example: make a QR attendance checker or a camera activity that detects movement for a science demonstration.

Webcam object detection, face/eye/smile detection, and motion detection demos.
QR scanners, attendance checkers, object-counting demos, and image processing tools.

6. 3D, AR, and game worlds

Example: generate a virtual gallery where students place artefacts, labels, audio notes, and reflection prompts.

3D object viewers, classrooms, galleries, games, physics simulations, and solar systems.
3D mazes, virtual museums, AR-style scenes, configurators, and 3D data visuals.

7. Games and activities

Example: turn revision questions into a timed quiz game with feedback, scoring, levels, and retry options.

Quiz, Snake quiz, drag-and-drop matching, physics puzzle, memory, and typing games.
Maths, programming, vocabulary, reward, and classroom learning games.

8. Chess and logic

Example: create a logic puzzle tutor that checks answers, gives hints, and explains the next best move.

Chess boards, puzzle trainers, best-move analysers, and beginner tutors.
Move legality checkers, play-against-computer apps, and notation viewers.

9. Maps and location

Example: build an interactive campus map for finding rooms, planning routes, and marking support services.

Campus maps, walking route planners, store locators, and distance calculators.
GeoJSON viewers, field data maps, and route or area measurement tools.

10. PDF and exports

Example: collect form responses and produce a polished certificate, feedback sheet, or summary report.

Certificate, PDF report, form-to-PDF, invoice, and feedback report generators.
Batch ZIP tools, Excel-to-PDF summaries, and screenshot-to-PDF apps.

11. Audio and animation

Example: design a pronunciation practice activity with sound, repeat attempts, progress, and celebrations.

Animated storytelling, soundboards, pronunciation practice, and audio quizzes.
Kids learning apps with sound effects, onboarding animations, and celebration screens.

12. Web app interfaces

Example: create a tidy project board for managing student tasks, deadlines, notes, and review status.

Admin dashboards, Kanban boards, task lists, galleries, modals, and wizards.
Mobile-friendly single-page tools with drag-and-drop and confirmation dialogs.

Best cool demo ideas

AI Prompt Builder plus live HTML preview.
Student CSV analytics dashboard.
Webcam motion detection app.
Diagram generator with PNG/PDF export.
Online colouring competition tool.
3D virtual gallery or interactive campus map.
QR attendance scanner, PDF certificate generator, or chess tutor.

Example prompt

All-in-one Student Learning App Generator

Build an all-in-one student learning app for a first-year class. Include a topic quiz, a diagram activity, a simple chart from sample class data, a reflection notes area, a PDF summary report, a QR attendance section, a short revision game, and a camera-based observation activity. Keep sample data local and make the interface suitable for classroom use.

Open TechTahi

This shows that TechTahi can create interactive educational tools from natural-language prompts, with examples, controls, outputs, and classroom-ready structure.

Thank
You

Any questions or suggestions?

Thank you for joining this workshop. If you have any questions, ideas, or suggestions, please feel free to get in touch.

Contact

Professor Minh Nguyen
ORCID: 0000-0002-2757-8350
Engineering, Computer and Mathematical Sciences
Phone: +64 9 921 9999 Ext.8214
Mobile: +64 21 175 4956

Email and Address

minh.nguyen@aut.ac.nz
mnguyen@aut.ac.nz
Auckland University of Technology
Computer and Information Sciences
St Paul Street, Auckland Central, Auckland 1010, New Zealand

Nga mihi nui. Thank you for your time, your ideas, and your participation.

Turn natural-language ideas into useful digital tools

Why this workshop matters

Who this workshop is for

Educators

AUT staff

Creative practitioners

Community innovators

What to expect

A short introduction to TechTahi and its purpose

A live demonstration

Guided hands-on activities using the system

Discussion of practical applications

Workshop outcomes

Practical applications

Education

Creativity

Community innovation

Guided hands-on activity

Step-by-step workshop task

Number -> Roman

Roman -> Number

TechTahi

What this 2-hour workshop is about

What the session includes

What participants will experience

What we will build and explore

Rationale: Why We Created TechTahi

Open to all

Built for safe experimentation

Low-cost digital wharenui

What this means in practice

First activity: get your first working code

Step-by-step guide

Use this first prompt

Prompt Builder text

What to check after Preview

If the first result is not quite right

Why the TechTahi workflow is practical

1. Choose from up to 8 LLMs

2. Use TechTahi as the free working interface

3. Reduce cost and setup overhead

4. Ready in seconds, not minutes or hours

5. Support semi-coding development

6. Save locally or share online

Main practical message

Second activity: improve a home loan calculator with Edit GUI

Suggested practice flow

Example prompt to start with

Prompt Builder text

What participants should edit

Why this matters

Risks and trade-offs of LLM use

What often happens in normal LLM use

Why that matters

What is deep learning?

What is an LLM and how does it run?

What are the costs and risks?

Take the good, prevent the bad

Harness the strengths of LLMs while reducing risks

No sensitive data leakage

Local execution after setup

Best of both worlds

Third activity: build a result announcement letter without sending sensitive data

Suggested practice flow

Example prompt to start with

Prompt Builder text

What participants should notice

Why this matters

New activity: local storage, encryption, and private file access

Activity 1: secured local diary

Prompt Builder text

Activity 2: local text file search system

Prompt Builder text

Why these examples are strong

Security and practical value

Next activity: improve ideas with Copy Enhance Query

Suggested practice flow

Starter prompts

Prompt Builder text

More complex system idea