Our First Portfolio Company

We started with 100 ideas and pain-points, distilled this down to about ten worth chasing, and further refined this down by validating four of these. In fact the original validated idea was applying AI to sole traders quote management and while this did resonate during testing when you hear “Yeah this is useful, but you know what is really a pain” your ears perk up.

Small and medium-sized businesses (SMBs) face a critical operational challenge, they miss significant revenue opportunities due to their inability to handle incoming inquiries while performing their core work. This problem creates a vicious cycle, when work volume is low, they can respond to inquiries, but as work increases, they start missing calls and messages, leading to inconsistent revenue streams and a yo-yo of work load.

Traditional solutions like hiring administrators are cost-prohibitive for most SMBs, with overhead costs easily exceeding $65,000 annually. The problem particularly affects trade services, contractors, and healthcare clinics, representing over 11.8M businesses in the US alone.

The Product

Jane provides an AI-powered virtual agent that autonomously handles incoming calls, scheduling and rescheduling, general inquiries, chasing bills and quoting, effectively breaking the missed-opportunity cycle and increasing the likelihood of customer conversion and satisfaction.

As we engaged customers on Jane, we noticed a strange phenomenon, we had various clinics using the product to handle their calls even for a product that was ‘mobile first’ and designed for trade services. This lead us to work closely with the clinic customers and recognised that there was also a strong vertical demand opportunity with this leading to the creation of Heron.

Key capabilities include:

• Autonomous call handling with business-specific knowledge

• Intelligent appointment booking and lead qualification

• Automated follow-ups and prioritization

• Comprehensive call summaries and lead quality assessment

• Web and mobile experience (desk bound and on the go workers)

• Integration with existing business tools and scheduling systems

• Connects with enterprise phone systems

• Frees up existing administration time

• Multi-lingual able to speak numbers languages including perfect English, Spanish, Mandarin and Portuguese

The Tech

Jane and Heron both gained a significant head start by taking advantage of the Ryft studio accelerators across agentic AI, backend, mobile and voice stack technology. Ryft has an agentic AI platform called Parrot which provides an easy way to voice and message enable existing products (the core I/O of any Agent). This allowed us to deliver incredibly low latency, global POP inbound numbers, session recording and analysis, interruption detection, noise cancellation and very low time to first token along with configurable models, multi-lingual speech providers (primary and fallback) etc.

The LLM model powering these products is customised to reduce hallucinations, stay on topic and steer people through a quick and concise engagement without wasting compute time and cost. Very important even with the significant price drop in tokens (10x reduction / year).

Tech is one thing, but making it accessible and simple to use is also a critical component of consumer facing products and this is where our experience shines. We made it as easy as downloading an app and a quick few questions and your own Jane is ready to work for you, handling your business 24/7.

Go To Market

Jane is a horizontal play across any SMB and sole trader as we see customers including builders, electricians, housing inspectors, cleaning companies, plumbers etc. We are targeting the US launch with the first funding after testing it in the AU/NZ market. There are approximately 33.2M+ small businesses in the US that would benefit from this type of product as they typically miss 62% of incoming calls and over 80% of callers will hang up on voicemail.

The channel and approach for Jane is through influencer marketing, online channels and product led growth. The lower price point means this is more a prosumer type play with self service and AI enabled customer support. Additionally there is an opportunity to do partnerships with platforms that Jane works with such as ServiceM8, Jobber, SimPro etc that are already in use by the target customer.

For Heron, which has a higher price point, there are approximately 1.8M clinics across the US. We see Heron being sold via more traditional approaches with outbound marketing, partnerships with existing PMS software and direct engagements with large clinic groups. There has already been interest here in AU/NZ and will follow a similar path to Jane of test locally and then push into the US.

All of these segments are suffering from increasing costs of staffing, difficulty with multiple languages, finding skilled staff and turnover. All these problems are quickly resolved with these products.

Competitors

There are existing competitors, some of these are behind “coming soon” signups and others are in market. This is to be expected in the arms race that is AI agent development. We see this as a sign of a healthy market and believe that vertical specialisation will be critical (therefore Heron) while also understanding that there are regional winners in the trade services SaaS tools space.

Some of the bigger risks would be “sherlocking" which is where an existing incumbent who provides a platform or service (ie: OpenAI) that is used by application developers suddenly absorbs the application functionality and offers it direct.

We believe the risk here is still able to be countered by building a strong brand, position and ease of use for specific customer niches and integrations.

We also believe the adoption curve of AI products with SMBs will take time to be realised, therefore leaning into the overarching venture studio portfolio balancing model of controlling costs (up/down the levers in engineering and marketing without layoffs and losing staff) to match timing without burning too much capital.

AKA venture strapping as we’ve discussed in previous posts.

Business Model

Both Jane and Heron are still experimenting with the business model. The good news is that there is one and currently it provides low churn, great margins and will only see improvements (to cost and performance) as the AI models continue to drop in price and increase in usefulness.

Jane has three tiers, with the solo plan being a three coffee / month type cost with limits on the number of calls and integrations, while the next tier steps up the integration (calendar, tools etc) and the number of calls and messages, then there is a bespoke business plan for higher capacity customers.

Heron is a simpler approach with a standard monthly (annual discount) price for a number of calls and inquiries with burst packs available automatically when volume goes over the tier limits.

It may be that pricing shifts to a price per resolution (as some AI companies are experimenting with, ie: Fin from Intercom) but currently we don’t see that as critical or necessary.

Pricing and the business model will shift as the capabilities of the agents increases and additional administration tasks are enabled.

Future

Crystal ball gazing in a world that is moving as fast as AI is going to look embarrassing when we re-read this in ten years time. Needless to say we expect to leverage the increasing intelligence in models and agents capabilities ie, autonomous browser usage instead of API calls for integrations, more thoughtful reasoning and action handling, less and less “human in the loop” required.

The key is gaining a foothold into the customers work experience, provide value today (as we do), and then continue to expand that value as AI enables us to eat more software and human workload leaving the SMB owners to focus on their unique skills and what they love and leave the administration to the agents.

We continue to believe in Jane and Heron as an agent of change for many small businesses and medical practices around the globe.