I’m very pleased to announce that Dan will be joining us next Monday. It has been discussed with Dan that he would work 24/7, 365 days a year. Dan will be ubiquitous, over our 18 offices over the world. Dan is the name of our new Artificial Intelligence (AI) company architecture.
Dan will be serving all of us and reporting to Sarah, who has been promoted to Chief Intelligence Officer. Sarah will remain at our Global Headquarters in Hong Kong.
This bold AI move has been requested by our board and shareholders. It will happen mostly through internal collaboration and external partnerships.
I strongly support the initiative and will personally follow its progress.
New job positions are offered at the end of this email. They reflect the transformation we are engaging in.
Dan will transform both back-office operations and customer-facing teams.
On back-office innovation:
No AI Department
No-one knows more than you which problems we are solving. We agreed that each function of the company will breed its own intelligence champions. Every department will benefit from Dan’s super-colleague capabilities. Last year, we cut IT costs by 15%. Some of these savings are dedicated to Dan’s AI expansion. Dan is maintaining the AI progress dashboard for our CFO in order for us to monitor each team’s pace on AI investments. We expect you to spend on AI.
AI Runs on Data
Last year our company’s annual report showed us again that trade execution can’t be a focus any more.
It took me a while to measure how critical data is and how functional algorithms are.
Customers and distributors come to us for our knowledge of their problems, their individual personalities and related solutions. Sarah has been missioned to increase our data production capacity, quality assurance and cyber-security layer around it.
To produce better data, user experience teams are missioned to rework tech architectures and user interface stacks to produce and capture more data and label it meaningfully. This better data will fuel Dan to provide you with better support, better investment opportunities and advice.
The fact that we serve our clients in 18 countries is one of our assets. We have the right data sets and now need to focus on labelling and tune them for the problems we are solving. Thanks to automation we shall see silo data issues puzzled out. Trade patterns from New York, London or Hong Kong will all be compared and shared live to all desks.
Don’t fear competition from internet giants, as their mountains of data do not relate to problems we are solving, nor does it create valuable advice to customers in the context of wealth growth and protection.
On the tech side, Dan will keep working on linear-regression-type AI methodology. This type of machine learning works on finding the line (hence “linear regression”) that goes as close as possible to as many data points as possible. When predicting Y in terms of X, Dan works on a 2D grid where data points are plotted in X and Y values. The linear regression finds the line as close to these values as possible. It mathematically takes a numeric error, and then minimizes that error value. This methodology fixes problems our customers ask us to solve; no reasons to defocus for now. Neural network capabilities will be built and tested before being integrated. A neural network is an example of what’s called a supervised learning algorithm.
Let’s say we don’t know if the equation is a line or not. As an alternative instruction to “here’s a line, minimize the error between it and the points”, we can ask to “make a prediction” – the feedforward phase – and “check the error to correct the predictor” – the feedback phase. Supervised learning means that we ask our model to make the prediction, then correct it when the prediction is off. That’s how we train Dan to get even better over time.
The Algorithm Design Team has been Downsized
We now mostly get algorithms from public domains or vendors, and craft them to our needs. We saw in the past months numerous acquisitions from our competitors. We all noticed that acquired companies did not represent fantastic technologies, so why this acquisition frenzy? It’s simple. Machine learning algorithms aren’t the secret sauce. The data is the secret sauce.
What does this mean? Wealth management leaders’ intellectual property and its competitive advantages are moving from their proprietary technology and algorithms to their proprietary data. Data becomes a more and more critical asset and algorithms less and less important. A lot of companies open source more and more of their algorithms. We’ll leverage that.
To all developers and curious minds, please note that we are flying you to Hong Kong for AI software training. Join those workshops and get rewarded in company currency tokens. Last month the internal company token marketplace best sellers were holiday trip bookings and upgrades of office computers.
When you start working with Dan, you’ll discover that we have started implementing new security features across the company. You won’t notice it, but we are constantly self-stressing our systems as well as evergreening our identifications system, computers and mobiles. I’d like to thank you for your constant cooperation on this matter.
Should you have any questions on the new security measures, Dan will answer all questions via @dan.
Dan joins us all for daily tasks: administrative queries, holiday applications, technical support, training requests, meal delivery and meeting arrangements. Dan will have a seat at all meetings as a diligent team member, always ready to trigger meaningful information and help with better decisions and work.
This gives me a perfect transition to the changes expected in our front-office operations:
Minimizing risk on all fronts and complying with regulatory initiatives has been and remains a priority.
Dan is a new colleague ready for heavy lifting and will be involved with client onboarding, know your customer (KYC) and anti-money laundering (AML). His toolbox integrates technologies from RegTech partners.
Our competitors are suffering on compliance. We bet on tech for cost reduction as agreed with shareholders. In addition, our compliance team will keep sourcing for new partners and technologies.
AI for Investment Performance
We’ve acquired a licence of proprietary data feeds from a quantum research company. This will feed Dan’s research capabilities and blend them with our own analytic sentiment and pattern capabilities. Expect unseen and high-value data to be extracted for our research and investment teams. Share those proudly with our customers.
While some of our investments will remain automated or indexed, we expect great performance again in actively managed fund teams with the help of Dan. I am a strong believer in collaboration between humans and machines. Our experimentations have shown that active managers get to outperform again with the help of AI tools. These headlines in the media on markets getting into the hands of passive indexing and algos won’t last. Active traders and their AI colleagues are coming back and we’ll lead this trend.
AI for Better Advice
On the advice side, our first beta test is very promising. It is crucial for our advisors to embrace the switch to assisted advice; we have restructured teams in that sense. Dan’s AI engine allows you to provide and communicate better portfolio decisions and tailored performance reports. Use the freed time to deepen relationships with clients and gather more assets.
No more late reporting any more, Dan will automate that through a live dashboard of customers’ interaction metrics and performance indicators.
Please come to the server room next Monday at 5pm to meet Dan and welcome him to our team.
You can reach Dan at:
- CRM & hotlines: @dan
- Trading & market terminals: @dan
- Cloud-based collaboration platform: @dan
New open job positions:
Data discovery engineer, institutional business
Small data hunter, you are working hand-in-hand with customers and investment teams to uncover quality data. You give a human feeling to our infrastructure so that customers are willing to open up to us more and unveil new qualitative data to deliver better services to them.
Algorithms buyer You work with products and investment teams to source and negotiate licences of the most useful algorithms from the public domain or marketplaces.
Data cleaning team lead You are passionate about data processing timelines and data deduplication, such as data compression techniques for eliminating duplicate copies of repeating data, patterns and outcomes finding. You help us to have high-quality data for our AI solutions.
Infrastructure and computer processing buyers You have experience in sourcing and assessing best-in-class cloud and computer processing services. Deep domain expertise in Oracle and blockchain required. Experience in distributed and parallel computing preferred.
Partnerships manager You will source and speed up the integration of technologies from external vendors.
Head of training You will design and execute a plan of training around our AI capabilities.
Our HR department has worked already with Dan’s talent-matching engine. If you haven’t been contacted yet or want to apply, message: @Dan_career
I want to thank our management team for embracing this exciting project and transformation. If you could see them in action like I do, you would know that they have remarkable capabilities, character, experience and wisdom.
Our company keeps learning and reinventing its game.
I’ve never been so excited.
Peter, CEO
Wealth Management Company PLC
Disclaimer: This is a work of fiction. The material here is not intended to provide, and should not be relied on for, advice. Names, characters, places and incidents are either products of the author’s imagination or are used fictitiously. Any resemblance to actual events or locales or persons, living or dead, is entirely coincidental. The author contributed this chapter in his personal capacity. The views expressed are his own and do not represent the views of his employer.