AI for Investors: How I Use AI to Make Better Investment Decisions
Every few months, someone in one of the investing forums I follow posts a question like: “Which AI company should I invest in?” That’s a valid question, but it misses something bigger.
AI isn’t just a sector to put money into. It’s already a working tool you can use right now to research deals, monitor your holdings, and cut through the noise. I’ve been investing in P2P lending, real estate crowdfunding, stocks, ETFs, and crypto since 2015, and in the last two years, AI has become part of my actual workflow in ways I didn’t expect.
This article is about the tool side of AI for investors, not the investment thesis. Here’s how I use it, and where it falls short.
Why Investors Should Pay Attention to AI as a Tool
Retail investors have always been at an information disadvantage. Institutional analysts have teams. We have evenings and weekends. AI doesn’t close that gap entirely, but it narrows it considerably.
The specific value for private investors is speed and synthesis. Reading a 180-page annual report, comparing the T&Cs of five P2P platforms, or staying on top of regulatory changes across multiple markets used to take hours. With AI, much of that becomes a 10-minute job.
This isn’t about replacing your judgment. It’s about freeing up your time so you’re spending it on judgment rather than on information processing.
Research and Due Diligence
This is where I get the most value. When I’m evaluating a new P2P platform or a real estate crowdfunding deal, there’s a lot of reading involved: the platform’s prospectus, their audited accounts, borrower terms, regulatory filings, and often a long thread of community forum posts where other investors share their experiences.
My current approach: I paste documents directly into Claude or ChatGPT and ask targeted questions. For a P2P platform, I’ll ask it to summarise the buyback guarantee structure, flag any carve-outs in the terms, and compare the stated default handling process against what I know from other platforms.
For annual reports, I ask it to pull out the key risk factors, summarise the loan book quality metrics, and highlight anything that’s changed compared to the previous year. What used to take a couple of hours takes 20 minutes, and I’m asking better follow-up questions because I’m not mentally exhausted from the initial read.
When researching platforms in my guide to European P2P lending, AI helped me process a large volume of platform documentation quickly and spot the structural differences that matter most to investors.
Portfolio Monitoring
My portfolio spans several platforms and asset classes. Staying on top of it manually is genuinely tedious. I now export data from the platforms that allow it, drop it into a spreadsheet, and use AI to help me write summaries and flag anything worth looking at.
Practically speaking, this looks like: pasting a table of returns and asking “what’s underperforming versus expected yield, and what’s the likely explanation?” Or uploading a PDF monthly statement and asking for a plain-English summary with any anomalies called out.
I also use it to draft the questions I should be asking. If a platform’s late loan percentage has crept up over three months, I might not know exactly what to look for in their next update. AI helps me construct the right questions before I go looking.
Risk Assessment
P2P platforms and real estate crowdfunding deals both involve risks that are genuinely hard to assess without legal or financial training. I’m not a lawyer. AI doesn’t make me one, but it helps me understand what I’m reading.
When I’m looking at a Spanish real estate crowdfunding deal, I’ll ask AI to explain the security structure in plain language: what does “first charge” mean here, what happens in insolvency, and what are the scenarios where I get nothing back. This won’t give me certainty, but it gives me a clearer mental model before I commit capital.
For P2P platforms specifically, I use AI to compare the risk disclosures side by side. When you’re evaluating five platforms, the differences in how they handle borrower default can be subtle and buried in dense legal language. AI surfaces those differences quickly.
My P2P lending guide and real estate crowdfunding overview both grew out of this kind of research process.
Tax and Admin
Investing across multiple platforms and countries creates a paperwork problem. Different platforms report income differently. Some send a consolidated statement; others make you piece it together from transaction exports. Several don’t send anything useful at all.
I use AI to help me understand what each platform’s tax documentation actually contains, identify gaps, and draft the questions I need to send to my accountant (or to the platform’s support team). It’s also useful for understanding the tax treatment of specific instruments. I’m not using AI to file my taxes, but it helps me go into those conversations with a much clearer picture of my own situation.
More practically: I’ve used it to help build spreadsheets that consolidate income across platforms, with formulas that handle different reporting formats. That kind of structured admin work is genuinely tedious without AI assistance.
Market Analysis and Staying Current
One of the hidden costs of being a serious retail investor is the reading load. Regulatory changes across the EU, rate movements, platform news, macro shifts that affect real estate yields. There’s more worth reading than any normal person has time for.
My approach is to use AI as a filter rather than a replacement for reading. I subscribe to several newsletters and research outputs, but I’ll often paste a long article or report into AI and ask for a two-paragraph summary with the key implications for retail investors in Europe. This lets me cover significantly more ground without just skimming everything.
I also use it to get up to speed quickly on topics I haven’t followed closely. If I want to understand what’s happened with ECSP regulation in the last 12 months, a good AI prompt gets me a useful starting point in two minutes instead of 30.
My Actual Workflow
To make this concrete, here’s what my AI use actually looks like week to week:
- New platform evaluation: I paste the key documents into Claude and run through a checklist I’ve developed over years of platform research. AI speeds up the extraction; the judgment about whether to invest is still mine.
- Monthly portfolio review: I export statements where I can, paste tables into AI, and ask for a summary of performance against expected yields and anything worth digging into.
- Deal-specific due diligence: For individual real estate crowdfunding deals, I use AI to parse the legal documents and project financials. I’ll ask it to stress-test the projections by changing key assumptions.
- Drafting questions: Before I contact a platform’s investor relations team or my accountant, I use AI to help me formulate the right questions. Better questions get better answers.
- Reading management: Any report or article over 1,500 words gets summarised before I decide whether to read the full version.
Most of this runs through Claude (my main tool) and occasionally through Perplexity for web-connected research on current events. I cover my full AI tool stack in a separate article if you want the technical setup.
What AI Cannot Do for Investors
This is important, so I’m not going to bury it.
AI cannot predict markets. It cannot tell you whether a P2P platform will survive the next credit cycle, whether a specific property deal will perform, or whether the macro environment will shift in a way that makes your current allocation look foolish. Anyone selling AI as a way to generate alpha is selling something I’d be sceptical of.
It also makes mistakes. When I use AI to analyse financial documents, I check anything that matters. It will occasionally misread a table, misinterpret a legal clause, or confidently state something that is simply wrong. The synthesis is useful; the outputs need verification.
What it cannot replace: the judgment that comes from having made and lost money in different market conditions. Knowing that a platform’s guarantees are only as good as the guarantor’s balance sheet isn’t something AI told me. That came from watching platforms fail.
AI makes me more efficient at processing information. It doesn’t make me a better investor in the sense that matters most, which is knowing when to act and when to stay still.
The Bigger Picture for Investors
If you’re already investing seriously, you’ll find AI most useful as a research and processing tool. The return on time investment is genuinely high for the tasks I’ve described above.
If you run a business as well as invest (I do both), the efficiency gains compound. The same tools that help me analyse a P2P platform also help me run AgentVania, where we build AI agents for businesses. Understanding these tools as a user has made me much more effective at deploying them professionally.
For a practical overview of how AI agents can work across your business and investing life, the guide to AI agents for business owners covers the foundations.
The investors who’ll get the most from AI in the next five years won’t be the ones who use it to make decisions. They’ll be the ones who use it to spend their time on the decisions only they can make.
I track the European P2P and real estate crowdfunding platforms I actually use in my P2P platform guide and real estate crowdfunding guide. Both are updated regularly based on my own portfolio experience.
Related
Leave a Reply