Some projects leave a mark well beyond their final deliverable. The Natural Hazards Technology Program (NHTP) grant project, run in collaboration with the Office of the NSW Chief Scientist and Engineer (OCSE), the Department of Customer Service (DCS), the NSW State Emergency Service (SES), and the NSW Rural Fire Service (RFS), is one of those.

Now that the project has wrapped up, we want to take a moment to reflect on what we built together, and to say a proper thank you to every partner who made it possible.

## **A Genuine Thank You**

Working alongside DCS, SES, and RFS was a privilege. These organisations deal with the highest-stakes situations imaginable, and they brought that same seriousness of purpose to every stage of our collaboration.

Their input shaped how we thought about translation at scale. Emergency communications need to be accurate, consistent, and produced quickly, often across many languages at once. The pressure of that use case pushed us to ask harder questions about our own platform.

We are grateful for the trust these agencies placed in VideoTranslatorAI. This project would not have been possible without their openness, their feedback, and their commitment to finding better solutions for the communities they serve.

## **What the Project Unlocked: Fragments**

The most significant product outcome to emerge from this collaboration is a feature we have named **Fragments**.

![Fragments for text translation](/api/media/elephantstripes-prod/cms/originals/blog-inline/1dea1086d11130eb7fafb1abf4407238305d5584c3fdcee0aea43817224e15c3.avif "Fragments for text translation")

At its core, Fragments is a sentence-level replacement tool for translation projects. It allows you to replace specific sentences or sections of text with pre-translated content, rather than sending the entire document through the AI translation engine every time.

Think of a large emergency briefing document. Certain sections are standardised. The safety disclaimers, the agency names, the procedural language: these phrases appear again and again across hundreds of documents. 

Without Fragments, the AI re-translates those sections every single time. With Fragments, you define those translations once, save them, and the engine pulls them in automatically wherever they appear.

Like glossaries, Fragments must be saved in a format compatible with the target AI translation engine, making them a structured and reliable part of your translation workflow.

## **Why Fragments Matter**

**Reducing redundancy in large-scale translation jobs**

When you are translating at volume, repetition is your biggest inefficiency. Fragments eliminate it. Recurring language is handled once, correctly, and reused across every relevant document.

**Saving real costs**

AI translation is token-based. Every word you send to the engine costs something. Re-translating the same sentence fifty times means paying for it fifty times. Fragments remove that waste, which adds up quickly on large or ongoing translation projects.

**Ensuring consistent, accurate translations**

Consistency is not just a quality standard; in emergency communications, it is a safety requirement. When the same phrase is translated by the AI in slightly different ways across different documents, it creates ambiguity. Fragments lock in the approved translation so the wording never drifts.

![](/api/media/elephantstripes-prod/cms/originals/blog-inline/05ca6a2ba260cbdb70cacd8c5b2606c6af91ef45fdb93631c7ae17442848e595.avif "Hybrid translation workflow by VideoTranslatorAI")

## **Looking Ahead**

We are proud of what this collaboration produced, and we are looking forward to seeing how Fragments performs across a much broader range of use cases beyond emergency services. 

Any organisation working with structured documents, compliance language, or high-volume translation workflows stands to benefit.

To OCSE, DCS, SES, and RFS: thank you. This one was worth every hour.

Wrapping Up the NHTP Project: Partners, Progress, and New Feature

How NHTP shaped one of our best text translation features Some projects leave a mark well beyond their final deliverable.

Wrapping Up the NHTP Project: Partners, Progress, and New Feature - VideoTranslatorA

Elephant Stripes builds applied AI products, internal tooling, and delivery systems.

Turn AI use cases into production workflows, internal tools, and customer-facing products with senior delivery judgment.

Practical AI Consulting That Ships

Elephant Stripes helps teams design, build, and ship practical AI products, workflow automation, and production-ready AI systems.

Elephant Stripes | Practical AI Consulting

Home

Elephant Stripes designs and ships practical AI systems for real business workflows.

VideoTranslator's efforts to build a solution in the emergency services space were recently profiled by [Asia Pacific Fire Magazine](https://apfmag.com/trusted-lingo-a-new-approach-to-cald-communications/), and we are very grateful for the opportunity to tell our story!

We'd love for you to check out the story, but here is a quick summary of the proposed 'Project Trusted Lingo'.

![Trusted Lingo envisages a three-step solution to inclusive communications at scale](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/6e65584a9a4e51262cd2084981a1d932721ae9c45c74c2014cf17f58b7e90a0d.webp)

Trusted Lingo is an innovative project aimed at addressing the communication challenges faced by culturally and linguistically diverse (CALD) communities during emergencies in Australia. The project proposes a custom AI solution that enables real-time translation of emergency broadcasts into over 20 languages. By bridging the language gap, Trusted Lingo aims to enhance the safety and understanding of CALD populations, ensuring inclusive and effective communication during crisis situations.

Effective communication is crucial during emergencies to protect lives and property. However, breakdowns in communication can lead to incomplete evacuations, trapped individuals, and even fatalities. Research shows that people with language barriers are slower to process messages not in their native language, making them particularly vulnerable during crises. In Australia, with approximately 7 million people speaking a language other than English at home, a significant portion of the population is at risk of failing to understand crucial instructions during natural hazards.

![Trusted Lingo's focal point is the effective communication to 'local information champions'](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/ac7134d8fa166194117f9c64d316c3c5c55afd6272817ca8387660f6f57c738b.webp)

Trusted Lingo aims to address this problem by providing a reliable solution that transcends language barriers. The project is a collaborative effort between VideoTranslator, Macquarie University, and the University of Newcastle. The proposed AI solution will offer real-time translation of audio, video, and text emergency broadcasts into multiple languages. It will utilize notification services to disseminate the translated content through different social media networks, targeting specific CALD communities.

The project's approach includes designating "local information champions" within the CALD communities, such as community organizations and social media groups. These champions will receive fully translated audio, video, and text communications minutes after emergency broadcasts occur, ensuring timely and accurate information reaches those who need it most. By involving local language-specific organizations, Trusted Lingo aims to enhance the effectiveness of emergency services' communication efforts.

To ensure the accuracy and reliability of translations, Trusted Lingo will develop glossaries of emergency services terminology, geographical names, and accurate translations. These glossaries will be used to test transcription and translation accuracy of emergency broadcasts, creating a library of reference material to train the AI continuously. The project also emphasizes the need to frame issues and solutions in a culturally appropriate way, understanding the differences between spoken and written languages.

In addition to translation capabilities, Trusted Lingo recognizes the importance of assessing the efficacy and engagement of end-users within CALD communities. The University of Newcastle's Lab for Applied Language Science will conduct research to evaluate the level of attention, comprehension, and digital trust among users of the proposed system. This assessment will provide valuable insights into the effectiveness of different communication methods, such as audio versus text, and guide future improvements.

![In development, what Trusted Lingo's interface will look like for triggering SMS notifications at scale.](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/3f9b096e1df1d2d0e43d994c72cc75680fadba9b027af702f1393cf4e5b9fb89.webp)

The success of Trusted Lingo relies on collaboration with emergency services and stakeholders in the field. The project invites expressions of interest from crisis communications practitioners, community engagement service members, and individuals engaged in coordination tasks during natural hazards. By involving these stakeholders, the project aims to ensure that the solution meets the rigorous requirements of emergency services and is tailored to their specific needs.

Trusted Lingo envisions a simple web application deployed at crisis communications centre's, enabling easy upload of audio or video files for automatic transcription and translation. The solution will generate multiple files with translated captions and the option for voiceovers, catering to different language needs. These language-specific videos will be automatically shared with pre-identified local information champions via social media channels, allowing for quick and widespread dissemination of critical information.

In conclusion, Trusted Lingo strives to create an accurate, cost-effective, and inclusive emergency communication system for today's diverse society. By leveraging AI technology and collaboration, it aims to break language barriers, ensure the safety of CALD populations, and improve overall emergency response and preparedness.

File Download: [Asia Pacific Magazine - Trusted Lingo.pdf](https://storage.googleapis.com/elephantstripes-prod/public/documents/APF_84_Banerjee_v2.pdf)

Trusted Lingo – A New Approach To CALD Communications

Trusted Lingo is an AI system that translates emergency broadcasts into 20+ languages instantly, ensuring everyone stays informed

DISCLAIMER: This blog post was updated after having a conversation with JobsForNSW. [3 Dec 2019]
**
Please check the JobsForNSW website for updates. The process covered below changes regularly and DO NOT rely on the below to be correct going forward. It was correct at time of publication, but talking to the JobsForNSW staff it is LIKELY to change pretty soon.**

Earlier this year we put in an application for a Minimum Viable Product Grant from JobsForNSW, a [NSW Government initiative](https://www.nsw.gov.au/).

In this blog post we are going to look at the entire process and work through our successful application. Specifically:

What is the MVP grant and where do you get more information?
What the grant process looks like and what information we had to provide to get the grant.
What the grant process was like, step by step, at a procedural level.

The point of this post is to provide information to anyone else within the Sydney/NSW startup ecosystem who is also looking into this grant and potentially applying.

Will probably do another post of the content of (a) the videotranslator.ai grant application itself, and (b) the progress report at the end.

We are also hoping to have a follow up post where we talk to one of the officers inside JobsForNSW to get a insider view into how the process works.

## What Is The JobsForNSW MVP (Minimum Viable Product) Grant?
From the JobsForNSW website, Minimum Viable Product (MVP) grants are designed to support promising startups with the funds needed to progress from a proof of concept stage to a minimum viable product. The grants are aimed at startups that are yet to generate revenue to help them engage with a potential business customer, or channel to market, and create innovative solutions that address compelling industry needs or market gaps.

## How Much Money Can I Get From The JobsForNSW MVP Grant?
From the JobsForNSW website, An MVP grant is a matched funding grant up to 50% of approved project costs, to a maximum of $25,000.

Remember: You do not have to ask for the full $25,000.

In the specific case of videotranslator.ai we had a total project cost of $41,400.00. The idea is JobsForNSW is funding 50% of approved project costs, to a maximum of $25,000. Hence the amount we ended up getting for our project was $41,400 * 0.5 = $20,700.

## Assuming We Get The Grant - What Is The Timeline For The Application Process?
From the JobsForNSW website, Successful grant applicants will receive 35% of the funding up-front and 65% after completion and validation of the MVP.

The exact sequence of events is provided below:

**DISCLAIMER: SmartyGrants below is the old system - JobsForNSW are transitioning to a new system.**
- You put in your application using an online system called SmartyGrants. The form itself is fairly straightforward - we did not use any external consultants.
- You get a notification saying your submission has been received. This is an email and looks like below. **Note the blacked out telephone numbers and emails are public, but have been hidden to the JobsForNSW staff don’t get spammed. If you really need to reach them, get creative with Google.**

![large_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_1_2d841642ba.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/b3b2d6cad41e3d63f80db42c944b867368afab8c72e923fcb555222e1d501099.webp)

- You get another notification saying you have been moved to round 1 of the selection process. Unless your application is terrible and you have not completed the form, you should get to here without too many issues. It looks like below:

![large_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_2_7333df57c9.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/94b2e99e9ab9dfd3f3359d084cd98de17d450b32322fbfe6bcc52df1b0895fec.webp)

- Once here JobsForNSW has an internal process. In another post we will talk about how that exactly works. For now, the application goes to an internal committee and either you get the grant or you don’t get the grant.
- In our specific case, videotranslator.ai got the grant - Winning! Also, on winning the grant you get notified by email. This looks like below:

![large_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_3_7d07efd6c7.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/540d11a63436a2bbb7028f119c2391d7db524193d9f85b7b2484602f2ffe2ede.webp)

- You will also get another email, from a person - that how you know :)

![medium_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_3b_a6b1b3e79a.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/7468d364e7740c615562758151a39f68e53b73ff73b07250cd2d0d835e9f2444.webp)

- Download the deed, and make sure you sign and return the contract.

![medium_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_3c_0908e6e6c5.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/27f7927293f5c4bd3b6b3bfff43e0e760ff1b293cc161421a983ad71b4571db5.webp)

- Sweet - the entire process took maybe 8 weeks - its a pretty quick turnaround.

## Assuming We Get The Grant - What Do I Need To Know About The Deed?
The grant is fairly simple. That being said, there are a few things which are really important:
- The grant will cover the budget you put into your application. You really want to get this part right!

![medium_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_5_6a2f148313.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/d1855567226fa0babac3abcda8b75eab14dfff12f7bb071ad87dcd799629ac1a.webp)

- Additionally, it is important that 80% of the Project Costs are within NSW. This does not mean if you pay Google Cloud or AWS or Azure you are in breach, it means that you can use a maximum of 20% of your budget on offshore resources. You also need to be aware that the duration of the projects is a maximum of 6 months.
- Don’t worry too much about being too specific - in private conversations, the JobsForNSW staff understand you are a startup and maybe not everything works out. They understand that you are trying to do something really hard, and not everything goes to plan always.

![medium_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_6_e082418728.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/081ad68120eec7ed87ca496fd3f1124151dd66530208c50b826eb13d37a160c1.webp)

- Lastly, the deed will cover exactly what JobsForNSW will fund and its timing. This looks like so:

![medium_blog_assets_winning_how_we_won_a_jobsfornsw_mvp_grant_winning_how_we_won_a_jobsfornsw_mvp_grant_7_281381b331.webp](https://storage.googleapis.com/vtai-web-media-prod/cms/originals/blog-inline/8674b32e896839333cf5acea1ccdffec7159dcc26754d266eeae30903df50c1d.webp)

- You're good to go!

## Assuming We Get The Grant - What Is The Timeline For The Project?
The main points to know:

1. The payment is in two parts. The first 35% is pretty much immediate. You raise and invoice and that is it.
2. The second part happens at the end of the project. You submit a report (using the same online application), have a meeting with your advisor, and then raise another invoice for the remaining 65% of the grant monies. 
3. That is kind of it in terms of process.

## Conclusion
We are super grateful to JobsForNSW. We did complete out prototype, filling out bits that did not work very well. We picked up a client (hopefully they will become a keystone client) during the project, and the client is super happy with our work.

Personally, we have not yet reached product-market fit, which I think of as having a scaleable sales process for your product at a good customer acquisition number (CAC), rather than product itself, but also think we are well on our way 😃 😃

## Thank You
I spoke to some other folks at Fishburners for advice through the process and after. Would like to thank the folks below, you guys are awesome.

1. David Cox at QAChef.
2. Man LEE at Sana.
3. Peter Boyer at Muster - a mate at Fishburners was curious about the MVP process. Hope you get the grant bud!

Check out the above startups - they are doing very cool technology stuff!

Please connect with us on LinkedIn, YouTube or Facebook for any comments, questions, or just to keep up to date with the work we do!

We are very grateful for your support!

Winning: We Won The JobsForNSW MVP Grant! Before, During And After...

We won the JobsForNSW MVP Grant! Learn about our journey through the application, execution, and completion of the grant in this behind-the-scenes blog.

We use cookies to improve site performance and understand how the portfolio content is being used.

Elephant Stripes is a consulting firm that designs, builds, and ships AI products and delivery systems.

Recognition media archive

Featured recognition coverage

Wrapping Up the NHTP Project: Partners, Progress, and New Feature

Latest recognition posts

Trusted Lingo – A New Approach To CALD Communications

Winning: We Won The JobsForNSW MVP Grant! Before, During And After...