In the latest episode of the NewIn Podcast, Cathi and co-host Hannah are joined by Hayley Millhouse (Co-Founder of Lumina Mortgages & Financial Planning) and Gary Abela (AI Strategy Partners).
What connects Hayley's story and Gary's expertise is a shared theme: the firms that thrive will be those who approach change with curiosity, build authentic relationships, and resist the urge to either over-automate or be paralysed by the options. Good advice, delivered well, is still the destination.
Hayley's route into financial services wasn't planned. Growing up in a household where her mum tracked every receipt and bank statement, she developed an early instinct for financial discipline. After work experience at a local building society, she landed a role at a high street bank and found she loved it.
Encouraged by colleagues to go to university, she later stepped into a private banking administration role, her first real taste of the regulated investment side of advice. Strong mentors and an appetite for opportunity during difficult times (including the 2008 financial crisis) helped her rise through the ranks, including a future leaders programme that gave her a seat in board meetings alongside area directors.
In 2016, her employer wound down the entire advice team. Hayley moved to a network, then joined a tech-led DA firm, gaining direct experience of working with the FCA. When that business wound up in late 2023, just after the birth of her second child, she called her friend of 25 years, Lisa, and together they decided to go for it.
"Would I have done it on my own? Probably not," Hayley admits. "It has been a real challenge."
From submitting their DA application to receiving authorisation took ten months. One significant setback: because Hayley hadn't personally advised on mortgages, the FCA initially refused to grant mortgage permissions for the new firm, a major blow to their business plan. Rather than give up, they engaged directly with their FCA case handler.
"If you don't ask, you don't get it. They are humans as well. As long as you're doing the right things and you can demonstrate that, you shouldn't have anything to worry about."
Once authorised, getting operational took longer than anticipated. Hayley names the tech stack as the biggest and most costly lesson learned.
"There are a lot of sales pitches, be mindful of that."
For a small firm, accessible customer service became a non-negotiable.
"We need to be able to pick up the phone and speak to someone."
She's encouraged by the direction of advice tech but realistic - there is still no unicorn that does everything.
Gary Abela has spent 25 years in financial services, from pre-RDR commission-based days through 15 years helping firms transition to fee-based models, and ultimately into AI, consulting for PE-backed consolidators on leveraging technology to solve the capacity problem.
Gary's core message: stop shopping for point solutions before you know what you actually want.
"It's very tempting to say 'I need a meeting notes solution, I need an automated suitability letter solution' and just go try those out. But you're not setting your growth strategy as a firm."
Map your operating model first, understand where the friction is, then evaluate vendors against that roadmap. That way you'll know whether they fit your journey, and where they don't, giving you a realistic view rather than being oversold because you're underprepared.
Newer DA firms don't have legacy tech baggage, but Gary is clear they still need a core back-office system. AI tools are currently automating pain points in workflows, not replacing the entire infrastructure. The advantage for new entrants is agility: watch how the landscape converges and be ready to switch if needed.
Gary is blunt about vendor roadmaps:
"Technology always takes longer and costs more money than people say."
His rule of thumb? If a vendor says three months, assume six. Many of these companies are under three years old, running flat out. Give them some grace, but build in exit clauses and be clear about what constitutes a loss of trust.
Gary cautions against applying large language models to the wrong use cases. For a complex pension consolidation, an LLM generating a suitability report will likely cost more time in human verification than if you'd written it yourself. He also flags a looming regulatory issue: as FCA suitability reviews evolve, firms will need to evidence why decisions were made. AI that generates plausible-sounding outputs without traceable reasoning may not survive that scrutiny.
Keep it simple. The cashflow tool in your back-office system might be good enough to start. Every extra integration creates complexity and data transfer risk.
Always ask vendors about data portability before signing: "If I decide I don't want to work with you anymore, how do I get my data back?" If the answer is vague, you could be locked in for a decade.
Hear the full conversation on the NewIn Podcast - available on Spotify, Apple Podcasts, YouTube, and Amazon Music.
The NewIn podcast is brought to you with thanks to Royal London, Lead Sponsor of the Verve Foundation.
