The risk labels on your investments may be wrong. Good Lioness ran the numbers

Brindha Gunasingham has spent more than 30 years in the investment industry, and one thing kept frustrating her: a lot of people were simply not in the best investments to reach their goals, and they didn't realise it.

The cause was hard to put a finger on, so she went digging. The digging turned into a PhD, and the PhD surfaced something fundamental: hidden assumptions sitting beneath most of the models used as the foundations of financial and investment theory.

She worked out how to solve one of those issues with an algorithm of her own. Good Lioness was born to commercialise it, using AI to make it available to everyone, and to help align investment decisions with real financial outcomes.

Brindha's analysis makes the stakes concrete. Looking at a 10-year investment in the US equity market, more than 84% of stocks would be misclassified by risk, placing them in the wrong risk band. When her team corrected for the hidden assumptions, more than 50% of those stocks moved two or three risk bands, potentially the difference between low risk and high risk.

The consequences cut both ways. Investors can be exposed to far more risk than they signed up for, or they can sit in investments more conservative than they need, quietly missing out on returns. Investment managers talk about the return they can get for you, she says, but rarely in enough detail about each individual's investment horizon, goals and constraints.

The product is an investment intelligence and decision platform. Its front door is Wealth Nexus, which starts by helping people understand where they are, because Brindha believes people need to understand what they're doing before anyone helps them do it.

The education is deliberately not a PhD thesis. The platform works out how each person likes to learn, whether that's small sound bites, bullet points, video or audio, and delivers the same underlying content in that form, matched to the goal at hand, such as saving for a house deposit or investing for their kids' education. AI works out the delivery; the substance stays constant.

From there, the platform unpacks each user's goals and how they view risk against each of them, then builds what Good Lioness calls the wealth genome: a dashboard showing where their wealth stands at any point in time, with tools to test out different scenarios.

For users ready to invest in listed equities, Good Lioness offers a large language model they can talk to. It tracks back to confirm which goal they're pursuing, then suggests the 10 most appropriate stocks for that scenario.

Brindha is emphatic about what sits underneath. This is financial advice, so everything is gated and locked down, and the conversational layer cannot wander off and advise something different. Beneath the LLM sit quant algorithms that are not AI-driven, built on 50 years of data and fully validated across eight equity markets.

She has a name for the architecture: an AI-enhanced model. In an industry where trust has been badly eroded, and where over 98% of people don't get financial advice because it's all too hard, that distinction is the product.

Brindha is not a tech founder in the usual sense. Her background is finance and investment, not IT, and she says the SMEC AI Pathfinder program made complicated ideas genuinely accessible: the nuts and bolts of how you build, the limitations of what AI can do, and the dangers of pure vibe coding without fundamentals underneath.

The program also covered the business side of launching a startup, and it changed her risk posture: it opened her eyes to security risks she needed to design for, something she calls not exciting, but important. Looking back, she says it accelerated her understanding across the board, sped up the build, and gave her confidence Good Lioness was on the right track.

Every week, the main session broke into small groups of AI founders comparing notes: their mistakes, their insights, their ideas. Brindha has since had detailed one-on-one discussions with several of them, solving problems in both directions. Guest speakers brought practical tips, and startup discounts on services are already proving handy.

The part she singles out is the weekly workshops: optional evening tutorials where participants could bring live build problems to the program mentor and each other, then go away and implement the answers the next day.

Her advice to founders weighing up the program is blunt.