Category Archives: Artificial Intelligence

Startmate Virtual Demo Day

Posted on by
Reply

Despite being under lock-down, the current cohort of Australian & NZ startups participating in the Startmate accelerator programme managed to deliver their Demo Day presentations on-line, including a virtual “after party” where founders were available for Q&A.

Given the large number of startups, and the fact that several were very early stage businesses, I have grouped them into loose clusters, with just a brief summary of each project. More info can be found at the links in the names:

Real Estate

Landlord Studio – tax & book-keeping solution for landlords. I tend to think the need for very niche accounting solutions is either overstated, or existing software platforms like Xero will come up with a plug-in of their own. Also, tax rules vary greatly by jurisdiction, so scaling internationally can be a challenge.

Passingdoor – an online estate agency trying to remove some of the costs and hassles of selling your home. Rather than listing with a traditional estate agent, Passingdoor will find buyers on your behalf (via a matching process?). I assume that prospective buyers will come from: people in the process of selling their own home; buyer advocates; or recent mortgage applicants – which is why the founders will need relationships with traditional agencies (referrals), mortgage brokers (cross-selling) and real estate ad platforms (leads). But given that sellers on Passingdoor only pay a 0.5% commission once an offer becomes unconditional, I’m not sure how the cashflow model will work.

MedTech

Mass Dynamics – scaling spectrometry for improved patient care. From what I understand, Mass Dynamics is using cloud-based architecture to “lease out” spectrometry capacity on demand, and to accelerate sample analysis.

LaserTrade – a marketplace for second-hand medical laser equipment. Rather than seeing re-usable equipment go to scrap, the founder saw an opportunity to create a marketplace for unwanted items. All items are tested beforehand. Has the potential to extend to other types of equipment, assuming the certification process is valid?

Health & Wellbeing

Body Guide – semi-customised rehab exercises to suit your symptoms. With superb timing as we emerge from months of inaction (or poor posture) while working from home during lock-down, this service is an aid to physical recovery, once your condition has been formally diagnosed. I’d probably want to check in with my GP or physio that the programme was right for me, though.

Sonnar – offers a library of audio content for people with reading disabilities. This is a subscription service, which claims to be cheaper than other audio-book services, and with a broader type of content (periodicals as well as books). I was unclear whether Sonnar is cheaper because they don’t need to pay publisher or author royalties (as it is deemed a charity?), or because they only license out-of-copyright content.

Good Thnx – promises to be “the world’s best gifting and recognition tool, with impact”. Aiming to provide a service for businesses, individuals and partner charities, Good Thnx is still in development. But as part of the Startmate Demo Day, gave attendees an opportunity to allocate a small financial donation to a selection of charities.

Food & Agriculture

Cass Materials – With the search for sustainable alternatives to meat, Cass Materials is developing a cheap and edible high fibre cell scaffold on which to grow cultivated meat – otherwise known as bacterial nanocellulose (BNC). I’m not opposed to the idea of “meat substitutes”, but I’m generally wary of what are sometimes called “fake meats” – vegetable proteins that are so processed so as to resemble animal flesh. I’d rather go vegetarian (I’m not sure I can go full vegan, because if we weren’t supposed to eat honey and yoghurt, why do they taste so good, especially together?).

Digital Agriculture Services – An AgTech platform is using AI-powered applications for developing a range of data-driven solutions across rural, agricultural and climate applications. The potential to bring more business insights and practical analysis to farming and allied industries is of huge potential in the Australian economy.

Heaps Normal – This company has taken a novel approach to producing non-alcoholic beer. Rather than chemical extraction or other processing to remove alcohol from ordinary beer, Heaps Normal has managed to brew beer without alcohol content.

Energy

Gridcognition – Using digital twin mapping of buildings, structures and locations to optimise the planning and operation of distributed energy projects. Given the value of lower transmission and storage costs, as well as more efficient energy generation, Gridcognition is aiming to bring their “decarbonised, decentralised, digitised” solutions to a range of industry participants.

ZeroJet – Helping the marine industry transition to sustainable energy solutions with the development of electric propulsion systems. In particular, targeting small inshore craft which are ideal boats for this type of engine.

Logistics & Analytics

PyperVision – This startup has developed a system for fog dispersal at airports. By aiming for zero fog delays, PyperVision is helping to reduce disruption in the travel and logistics sectors.

Arlula – An API service to stream satellite images from space. As we know, satellite imagery is an important input to modelling, planning and analysis. Arlula also offers access to historic and archive content.

Database CI – A platform for in-house software developers to access the right sort of enterprise data for real-life testing purposes. For example, realistic and appropriate “dummy” data that does not compromise privacy, confidentiality or other obligations.

Law on Earth – On-line access to self-serve legal documents, forms and precedents, plus lower-cost legal advice. With a mission to “empower the public to safely manage their own legal needs”, Law on Earth already has a tie-up with Thomson Reuters, one of the largest legal information providers in the world.

Next week: Are we there yet?

The Limits of Technology

Posted on by
Reply

As part of my home entertainment during lock-down, I have been enjoying a series of Web TV programmes called This Is Imminent hosted by Simon Waller, and whose broad theme asks “how are we learning to live with new technology?” – in short, the good, the bad and the ugly of AI, robotics, computers, productivity tools etc.

Niska robots are designed to serve ice cream…. image sourced from Weekend Notes

Despite the challenges of Zoom overload, choked internet capacity, and constant screen-time, the lock-down has shown how reliant we are upon tech for communications, e-commerce, streaming services and working from home. Without them, many of us would not have been able to cope with the restrictions imposed by the pandemic.

The value of Simon’s interactive webinars is two-fold – as the audience, we get to hear from experts in their respective fields, and gain exposure to new ideas; and we have the opportunity to explore ways in which technology impacts our own lives and experience – and in a totally non-judgmental way. What’s particularly interesting is the non-binary nature of the discussion. It’s not “this tech good, that tech bad”, nor is it about taking absolute positions – it thrives in the margins and in the grey areas, where we are uncertain, unsure, or just undecided.

In parallel with these programmes, I have been reading a number of novels that discuss different aspects of AI. These books seem to be both enamoured with, and in awe of, the potential of AI – William Gibson’s “Agency”, Ian McEwan’s “Machines Like Me”, and Jeanette Winterson’s “Frankissstein” – although they take quite different approaches to the pros and cons of the subject and the technology itself. (When added to my recent reading list of Jonathan Coe’s “Middle England” and John Lanchester’s “The Wall”, you can see what fun and games I’m having during lock-down….)

What this viewing and reading suggests to me is that we quickly run into the limitations of any new technology. Either it never delivers what it promises, or we become bored with it. We over-invest and place too much hope in it, then take it for granted (or worse, come to resent it). What the above novelists identify is our inability to trust ourselves when confronted with the opportunity for human advancement. Largely because the same leaps in technology also induce existential angst or challenge our very existence itself – not least because they are highly disruptive as well as innovative.

On the other hand, despite a general shift towards open source protocols and platforms, we still see age-old format wars whenever any new tech comes along. For example, this means most apps lack interoperability, tying us into rigid and vertically integrated ecosystems. The plethora of apps launched for mobile devices can mean premature obsolescence (built-in or otherwise), as developers can’t be bothered to maintain and upgrade them (or the app stores focus on the more popular products, and gradually weed out anything that doesn’t fit their distribution model or operating system). Worse, newer apps are not retrofitted to run on older platforms, or older software programs and content suffer digital decay and degradation. (Developers will also tell you about tech debt – the eventual higher costs of upgrading products that were built using “quick and cheap” short-term solutions, rather than taking a longer-term perspective.)

Consequently, new technology tends to over-engineer a solution, or create niche, hard-coded products (robots serving ice cream?). In the former, it can make existing tasks even harder; in the latter, it can create tech dead ends and generate waste. Rather than aiming for giant leaps forward within narrow applications, perhaps we need more modular and accretive solutions that are adaptable, interchangeable, easier to maintain, and cheaper to upgrade.

Next week: Distractions during Lock-down

 

 

 

 

 

 

Pitch X – Launch Into A New Decade

Posted on by
Reply

Last week I was invited to be one of the judges at the final Pitch X event of 2019 (and of this decade), organised by Academy Xi and Melbourne Silicon Beach Group, and hosted by YBF Ventures. My fellow judges were Abena Ofori of MAP and Michelle Bourke of Foresight Digital.

As usual with Pitch X, each pitch was given 90 seconds to present, and the top 3 were then invited back for a 5 minute pitch. After each pitch, the panel of judges were given time for Q&A. The pitches in order of presentations in the first round were:

MotionAI

Remote monitoring system for people who require care, in case they fall or need assistance in their home. Designed around a combination of machine learning, AI and motion sensors (that don’t track facial recognition). Only decision-based information is sent to the monitoring network.

Sola.io

Investment platform to fund solar power under a virtual power plant structure, bringing together investors and producers, who might not otherwise have access to the financial and production benefits of this renewable energy resource.

Oyumz

Bringing home-cooked meals to the food delivery market. Currently in beta launch, looking to bring on new cooks and suppliers. Limited number of providers at this stage, and having to manage regulatory (food hygiene and licensing) and logistical (delivery, inventory, geography) challenges.

CPAP Buddy

Developing CPAP (Continuous positive airway pressure) oxygen masks, designed for premature babies, and intended to prevent brain damage or other injury that can arise from incorrectly fitted or poorly designed devices. Combines real-time monitoring with continuous visual feedback and detection of interfacial contact pressure.

Travels by TM

Helping people to gain the confidence and resilience to go travelling alone. Part curated travel planner, part counselling course, part self-help guide. As judges, we felt it was difficult to see how this business would scale, given the very personal nature of the service.

Mentor Community

Positioned as a match-making mentoring platform, it is designed to overcome some of the challenges people can experience in trying to find a suitable or appropriate mentor. Very difficult to know what technology is being deployed (to match mentors and mentees), in what is an uneven “market place” – more people seeking mentors than there are people willing or able to mentor them. And no opportunity to examine the financial model.

Fulfilled

Bringing “zero waste” cleaning products to the market. Distributes concentrates, that simply need diluting in water, and avoids the use of single-use packaging. Using Australian-produced botanical ingredients. There was some confusion on the business model – the pitch mentioned a home delivery subscription service, and supplying to professional cleaning companies.

VibeDate

Describing itself as a curated service offering unique and personalised dating experiences. No doubt there is a market for outsourcing your date decisions (or just to get some fresh ideas), but this was another pitch that would be difficult to scale, and again, it wasn’t clear how technology is being deployed in the solution.

PetMate

A total marketplace for pet products, services and solutions that also uses ML, AI and data analytics to track, recommend and predict your pets’ needs.

Cari

Another customised CPAP solution for premature babies, but also targeting neo-natal infants with sleep apnea conditions. Already at prototype stage (and scoping manufacturing options for medical grade silicon), but with at least 2-3 years of clinical trials before achieving medical device approval, the team have already identified multiple channels to market including hospital purchasing committees.

The three winning finalists were 1st: Sola.io; 2nd: MotionAI; and 3rd: Cari

As part of their prize, the winners will be featured in this blog in the near future. Stay tuned for more updates…

Next week: Signing off for Saturnalia

Startup Victoria – Best of the Startup State Pitch Night

Posted on by
Reply

In support of Victoria’s reputation as “Australia’s Startup State”, last week’s Startup Victoria pitch night was designed to showcase four of the best local startups. Hosted by Stone & Chalk, the judges were drawn from Mentorloop, Brosa, Giant Leap Fund, Rampersand and Vinomofo.

The pitches in order of presentation were (website links embedded in the titles):

Code Like A Girl

Founded four years ago, Code Like A Girl’s stated mission is to bring greater gender diversity to the ICT sector (information and communications technology), within both the industry and education spheres. To do this, the founders say we need more female coders, which they plan to achieve via coding camps, internships, and community events. Positioning itself as a social impact enterprise, the business is active in four States, and 75% of interns are placed into full time roles.

To support the ongoing development of its “role ready” value chain and to prepare for possible overseas expansion, Code Like A Girl is seeking $1.5m in seed funding. Currently piloting the training model via education providers (RTOs, boot camps, universities, online code schools), the business takes a 10% commission on courses sold (held twice a year), plus it charges placement fees of $2k per person.

But the model is difficult to scale, especially as Code Like A Girl does not own or create the actual training content – it is acting as a sales channel for third party courseware, and providing platform for advocacy, engagement and influence. Its key metrics are based on things like social impact scores – such as 30% of kids return to boot camps. The panel felt that the community platform is a huge cost centre, and it might be preferable to try a TedX model, where Code Like A Girl provides branding and foundational support to build more of a network effect – but without its own curriculum, the business will still struggle to scale.

Seer Medical

The business claims to make epilepsy diagnosis easier, and is currently raising $14m for European expansion (UK & Germany). To improve current diagnosis, the model needs to capture time series data to distinguish epilepsy from other conditions, but do so faster, cheaper and more efficiently than current processes. Founded in 2017, Seer has already serviced more than 1500 patients via 200 clinicians.

Using the Seer Cloud infrastructure,  it can achieve diagnostic outcomes 10x faster than traditional methods, and the platform is using machine learning to train its algorithms. The service is subject to Medicare reimbursement, which has no doubt assisted adoption.

Asked by the judges if the platform could be used to diagnose other conditions, the founders mentioned cardio, sleep and other health domains. As for competition, this comes mainly from the status quo – i.e., hospital based services. With advocacy from neurologists, giving them access to customers, the founders have a strong track record in the research field, which helps to open doors with clinicians. Along with research partnerships, plus the public health cost reimbursement, data is the fuel of the business –  Seer even have access to some third party data on which to train their diagnostic.

Liven

A dining rewards app, Liven is also bringing a behavioral gamification layer to a real world use case. Currently, there is a poor linkage between loyalty programmes and gamification. So, Liven has launched a universal reward token (the LVN token) for use in a digital/real world context.  The details were scant, and the status of the LVN token sale is unclear, but it seems users can earn LVN tokens from completing certain “missions”. The token (using a standard ERC 20 token format on the Ethereum blockchain), is designed to be interoperable and fungible (but Liven does not yet appear to use blockchain in its end user app or merchant point of sale solution).

The said merchants pay a 10-25% commission on app-based sales, of which upto 40% is paid back to the end user in the form of LVN tokens – if I got the maths right, Liven itself is securing $15 profit on every $100 of sales. Currently only available in Melbourne and Sydney, the judges wanted to know what the appeal is to merchants. According to the founders, users typically spend more in an average transaction when they use the app. It also seems that the app only works in brick and mortar restaurants, cafes and bars. The path to scaling will be via channel partners such as PoS systems.

Although not yet deployed, in future, it was suggested that users will be able to pay in any crypto – which raises all sorts of questions about the tokenomics of the LVN token, and whether LVN will be subject to exchange rate volatility (and even token speculation) or act as a stable coin; if the latter, what will it be backed by or pegged to?

Phoria

Phoria is in the business of extended reality technology (XR). Started in 2014, Phoria was an entrant to the Melbourne Accelerator Programme (MAP), with the stated goal of moving VR into a mobile experience (“democratize VR”).  Having gained some clinical VR research experience, Phoria has since worked on commercial projects such as “Captured” (turning a 3D scan of a building or structure into a Digital Twin), “Rewild Our Planet” (a Singapore-based AR experience), and various art installations museum exhibits.

Phoria is commissioned by tech and media brands to create XR content. It has developed a SaaS model, whereby it can turn real space into virtual space (“virtualising internal space”).

The judges wondered where we are along the cycle of mass adoption vs peak hype. In response, the founders mentioned that the first wireless headsets are now available, although consumer-facing mixed reality hardware is still 3-5 years away. With a growing customer base in engineering and architecture applications, Phoria’s main focus is on spatial information.

After the votes were counted, the People’s choice was Seer Medical, who also won the overall prize.

Next week: 30 years in publishing