This September brought a wave of notable AI advancements that impact finance.

There’s plenty to catch up on!

Here’s a quick look at this month’s highlights:

1. Microsoft 365 Copilot Wave 2: Excel Enhanced with Python
2. OpenAI Launches o1 Model Series, Enhancing Problem-Solving
3. Google Introduces NotebookLM: The Lifesaver You Didn’t Know
4. Oracle AI Agents: Elevating Finance with Automation and Insight
5. UBS Launches AI-Powered M&A Tool for Rapid Deal Analysis

1. Microsoft 365 Copilot Wave 2: Excel Enhanced with Python

Microsoft’s latest Copilot update brings Python into Excel. Users can now work with Python’s libraries, like Pandas and Matplotlib, through simple, natural language commands within Excel. This allows for advanced analytics, forecasting, and data visualization without leaving the spreadsheet.

For finance professionals, this means the ability to handle complex tasks like data cleaning and machine learning within Excel. This update boosts Excel’s role in financial analysis and decision-making.

My take:

If you often switch between Excel and Python, this update simplifies your process. It removes the need to move data between tools, making advanced analysis more accessible.

Financial analysts and data scientists can now automate, visualize, and analyze data in one place. Generating Python code with easy commands will make many daily finance tasks simpler and faster.

Interesting articles on the topic:

• “Anaconda Announces General Availability of Python in Excel, Transforming Data Analysis for Millions of Excel Users​” by Inside AI News
• “Python in Excel – Available Now” by Microsoft

2. OpenAI Launches o1 Model Series, Enhancing Problem-Solving

OpenAI has introduced its latest AI models, known as OpenAI o1, which were previously code-named “Strawberry.” These models are built for advanced reasoning, and designed to handle complex challenges in fields like science, mathematics, and programming. What sets the o1 models apart is their ability to “think” before responding, mimicking human thought processes. By breaking problems into smaller steps, they improve accuracy in areas like physics and coding.

The o1-preview model has already shown strong results, outperforming GPT-4 in tasks that demand deep reasoning, such as debugging and competitive programming. OpenAI also released o1-mini, a smaller, more affordable version, to make these capabilities available to more users.

My take:

This release could change how we tackle problem-solving in fields requiring high-level reasoning. The ability of o1 to work through multi-step problems and correct itself is a major advancement.

In finance and business, this could lead to quicker, more reliable solutions for complex challenges.

Last week I was training the team of Hashicorp which is in the process of getting acquired by IBM. I used o1 to perform a gap analysis on their accounting policies and those of IBM.
It took me 2 minutes to come up with the recommendations and area of investigation.

Imagine doing this task yourself! It would have probably taken a few days. Check the video here.​

Interesting articles and links on the topic:

• “Use o1 for month-end accounting tasks” by Jason Staats​
• “OpenAI Releases o1, a Model With Reasoning Abilities” by CDO Trends​

​

3. Google Introduces NotebookLM: The Lifesaver You Didn’t Know

Google has introduced NotebookLM, an AI-powered tool designed to help users quickly synthesize information and generate insights. Originally known as Project Tailwind, this experimental assistant allows users to upload their own sources, like Google Docs, and receive summaries, topic suggestions, and follow-up questions from the AI.

What sets NotebookLM apart is its “source-grounded” approach. It bases responses directly on the content provided, ensuring accuracy and relevance. This makes it especially useful for researchers, students, and professionals working with complex material. Whether you’re trying to understand detailed reports or brainstorm ideas, NotebookLM simplifies the process. ​

My take:

NotebookLM has the potential to become a go-to tool, though it hasn’t received much attention yet. Extracting key figures and spotting trends can be time-consuming, but this tool automates the tedious parts. It allows users to ask specific questions about their own documents and get fast, targeted insights. This could save time and surface important details that might otherwise be overlooked.

Did you know NotebookLM can turn your uploaded documents into a podcast, so you can listen to them anytime, anywhere? Yes, your documents become a podcast.

For a detailed look at NotebookLM, here is a video from Alex Banks and Christian Martinez also did a demo of this tool.

Interesting article on the topic:

• “Introducing NotebookLM” by Google
• “Google Introduces NotebookLM: Your Personalized Virtual Research Assistant” by Analytics Vidhya

4. Oracle AI Agents: Elevating Finance with Automation and Insight

Oracle has introduced over 50 AI-powered agents embedded in its Fusion Cloud Applications Suite. These agents help automate business processes and provide personalized insights across functions like finance, HR, supply chain, sales, and marketing.

By handling repetitive tasks, they allow employees and managers to focus on more strategic work. For example, they can manage scheduling, candidate sourcing, benefit analysis, and customer service improvements.

Oracle’s AI agents also use predictive models and natural language processing to improve decision-making. They offer real-time insights and recommendations tailored to specific roles, aiming to boost efficiency and support organizational growth. (

My take:

Having agents available on top of our tool stack is what I am looking for the most in the advancement of AI for Business and Finance. From the list, I like the anomaly detection in ledgers and the agent on forecasting as they can reduce human error and increase efficiency.

Interesting articles on the topic:

• “Oracle AI Agents Help Organizations Achieve New Levels of Productivity” by Oracle
• “Oracle Adds Powerful New AI Capabilities Across Oracle Fusion Cloud Applications Suite” by Clover Infotech

5. UBS Launches AI-Powered M&A Tool for Rapid Deal Analysis

UBS has introduced a new AI tool designed to transform the mergers and acquisitions (M&A) process by analyzing over 300,000 companies in less than 20 seconds. Known as the M&A co-pilot, this tool helps UBS quickly identify buy-side and sell-side opportunities.

It goes beyond basic analysis, offering insights into activist campaign targets by examining the tone of management communication during presentations and Q&A sessions. This positions UBS at the forefront of using AI in M&A, giving it a significant advantage in delivering faster, more accurate services.

My take:

For financial institutions and deal-makers, speed is crucial when identifying M&A opportunities. UBS’s AI tool cuts the time needed for complex analyses, offering near-instant insights that could change how M&A transactions are done.

In the coming weeks, we will cover how to use AI for M&A in one of our articles.

Stay tuned!

Interesting articles and links on the topic:

• “UBS Unveils AI-Powered M&A Tool That Analyzes 300,000 Firms in 20 Seconds” by AI Insights​
• “AI-Powered M&A: UBS Tool Analyzes 300K Companies in Seconds” by Techopedia

That wraps up an exciting month!

If you come across any interesting articles or insights, feel free to share them in our Slack channel.

I hope this digest helped you find some valuable takeaways.

Let’s keep the conversation going and continue learning together!