Key considerations for modernizing asset management in the AI era

For decades, leading trading desks won through technological craftsmanship that delivered an “artisanal edge.” Custom stacks, local compute clusters, and bespoke data feeds that once set firms apart now create friction and operational burden. While industries from Big Tech to banking shifted to platform-based model environments, many trading desks and asset managers have hesitated. The result is that margins are being negatively impacted and the pressure is on to change. The need to rapidly adopt AI only heightens the urgency. Legacy infrastructure and fragmented workflows increasingly constrain progress. A unified platform is not just an essential, but a strategic must-have.

McKinsey’s Asset Management 2025: The Great Convergence and North American Asset Management 2024: Beyond the Balance Sheet reports highlight the same challenge confronting trading desks still operating in artisanal, fragmented environments: while assets under management have surged to record highs, costs are rising even faster because they are being driven by operational complexity, legacy infrastructure, and what McKinsey describes as a “compounding operational debt” across the industry. This creates a "vicious cycle," where firms are forced to spend their technology budget on "run-the-business" maintenance, leaving as little available for the innovation required to compete.

The problem is exacerbated when AI is added to the mix. According to a 2025 enterprise survey, nearly four-in-five firms report difficulties integrating AI with their existing systems, and 35% of AI-leaders cite infrastructure integration as the top barrier. Cognazant describes legacy infrastructure as incompatible with AI’s rapid pace. For trading desks still operating on fragmented or bespoke tech stacks, these findings present a clear warning: without modernization, AI ambitions may remain aspirational.

Not only are there cost and scale issues, the artisanal edge has vanished. Python is the lingua franca for quantitative finance. Cloud computing allows standardized environments and eliminates local compute constraints. Data marketplaces made once-exclusive datasets broadly accessible. Open source models now run easily on GPUs. The tech stack that used to distinguish firms has become commoditized so firms need to find a new way to differentiate themselves. What’s needed is a shift to a unified, collaborative, and governed platform that allows firms to rewire their trading engines for speed and scale.

Top three benefits of AI platforms for asset management

Unified AI platforms provide three enormous benefits:

1. Research-to-trade cycle-time collapse

Many desks suffer a value-crushing time lag between research and production. Model builders hand off code to validation teams who rewrite it for testing; after back-and-forth delays, it moves to technology teams for production and finally becomes available to use in trading.

A KDnuggets industry poll revealed that 80% of machine learning models stall before deployment due to these fragmented workflows. Undeployed models are worse than worthless. They create negative ROI because cost is incurred but no value has been delivered. Unified environments set value free.

• Moody's reported that by industrializing their platform, they achieved 50% faster model development and were able to deploy models 6x faster.
• Even more impressively, a Large Canadian Bank demonstrated that a unified platform increased their research-to-production velocity by as much as 77%.

Reducing research-to-trade cycle time lets quants and traders deliver value from insight at the speed of the market.

2. Model governance at market speed

Global standards like SR 11-7, MiFID II’s RTS 6, and the Fundamental Review of the Trading Book (FRTB) require transparency and strict controls. Traditional approaches treat model governance as an afterthought creating delays and exposure, but unified systems optimize risk management speed and control. Validation becomes continuous: models, data, and results are automatically registered, lineage is tracked, and evidence is generated in real time.

UBS integrated governance directly into their modeling workflows using their unified platform. They reported strengthened regulatory compliance and reduced the time to bring models to production.

3. Tech debt optimization

Each "special" trading environment that once created an edge now adds cost and complexity. What was once an advantage is now a hindrance to value creation. A 2024 Accenture study found that financial institutions that consolidate their tech stacks achieved 40% higher profits and 60% higher growth rate than their peers.

In addition, research shows that industrializing model lifecycle operations through a unified platform can reduce cost bases by 25 to 40 percent, turning technical debt from an unbounded liability into a controlled variable.

Why past performance does not guarantee future results

When contemplating these changes, the question every CTO and head of trading asks is: If everyone uses the same platform, where's my competitive advantage? But, trying to leverage past methods does not guarantee future success. Sydney Finkelstein warns in Why Smart Executives Fail, firms that do that are clinging to “yesterday's answer”; a reality that once worked but is no longer valid.

Leading firms are realizing that their edge comes from strategy development and idea-to-execution speed. Adopting a unified, governed, and scalable platform is the key to enabling competitive advantage and future-proofing your environment.

Key asset management success stories

Global bank: A Large Canadian Bank faced the challenge of rapidly scaling model-driven services. After adopting a unified platform, they reduced the end-to-end time to build and deploy new models by 77%, while achieving 6x faster model development. The result was a $175M increase in gross margin and $7.5M in IT and data science cost savings.

Hedge fund: Coatue, managing $15B in assets, needed to rapidly identify signals to inform investment decisions. After adopting a unified platform, they reported "orders of magnitude productivity gains" by allowing experiments to run in parallel. Their Head of Data Science noted that iterating on ideas faster speeds up the research process, allowing them to "rapidly improve and deploy new strategies".

Asset managers: Man Group, a pioneer in quantitative investing, industrialized their research platform to flip the script on productivity. They reported that while researchers historically spent "80% of their time coding and 20% thinking," their modern platform reversed this ratio. Research tasks that once took "hours or days to code and debug" are now accomplished in "minutes," allowing them to scale alpha generation without linearly scaling headcount.

Future-proof your asset management platform strategy to ensure AI readiness

The transformation to a unified platform is a journey that requires aligning people, processes, and technology. Leading firms have already started their journey because they know the cost of delay is far higher than the price of change.

To estimate the value of this approach for your company, calculate your velocity tax. Measure how many days it takes to move a trading idea to a working model, validate and approve the model, move it to production and make a trade. If it's more than two weeks, you're falling behind. Now calculate your technical debt tax: how many engineering hours per month are spent maintaining custom pipelines versus the hours it would take to manage a unified platform? As a benchmark, we’ve seen up to 70% reduction in those hours.

The firms that win the next decade won't be the ones with the cleverest tech stacks. The ones with the most efficient industrial machinery will dominate. The question is not whether to modernize your trading desk, but whether you can afford the cost of remaining an artisan in an industrial age. To gain further financial services insights, check out the data science and AI leader playbook. Leaders who are at the top of their game share proven strategies on how to scale people, processes, and technology in today’s evolving landscape.