3 ms·
Ask HN: Looking for long-term investors to test tool to reduce overtrading I’m building a tool called InvestAssist. It started as something I built for myself.
by amykummetha 9mo ago
Ask HN: Looking for long-term investors to test tool to reduce overtrading
I’m building a tool called InvestAssist. It started as something I built for myself.
I enjoy researching companies and deciding what I want to own long term. What I struggle with is execution. When to buy. When to wait. When to do nothing. I found myself re-checking charts every morning and second-guessing decisions.
So I built a daily pre-market brief that:
- Analyzes only the assets on your personal watchlist
- Runs predefined quantitative checks (momentum shifts, pullbacks, stabilization, trend state)
- Classifies each asset into a clear condition
- Explicitly tells you when no action is warranted
It’s not automated trading. It's not personalised investment advice.
It’s a structured monitoring and analysing layer to reduce manual decision loops, overtrading and emotional trading. That's why its tag line is 'Patience-as-a-Service'.
The idea is that if you're going to time the market anyway, would you rather wing it, or do it manually and repeatedly every morning, or automate the entire workflow? InvestAssist is designed for the latter.
It ingests historical OHLVC data via a third-party API, processes indicators in a serverless AWS pipeline (Lambda + Step Functions orchestrated by EventBridge), stores state in S3, and renders a per-user brief each morning before market open.
I’m running a 90-day pilot and looking for ~30–40 long-term investors (stocks or ETFs, not day traders) to see if it does reduce time investors spent making decision, decision fatigue and implusive trading. If you’re interested, please check out invest-assist.com.
I know that asking for commitment for 90 days is a lot, so as a token of appreication from me, I'll give the participant who completes the pilot a 50 eur gift card.
Happy to answer any (technical) questions you have.
- andsoitis 9mo agoWhat elements, besides price, signals a buy recommendation in your tool? Asking because if you’re a long-term investor and specifically interested in a particular company, you can already setup price alerts or limit or stop orders. Presumably you’ve already done fundamentals research on the equity and so don’t need daily eval.
- amykummetha 9mo agoYes, it’s purely price-derived. I’m not doing fundamentals inside the model. The assumption is you’ve already decided what you want to own long term. The difference from a simple price alert is context and adaptability. A basic alert might trigger at “-10%.” But that’s static. It doesn’t adapt to the asset or the environment. If the stock is still in a sharp downtrend, that alert alone doesn’t tell you much — you might just be buying into a falling knife. The tool evaluates multiple price conditions together: short- and medium-term momentum, trend direction, where price sits within its recent range, and recent volatility. It only flags a buy-type condition when there’s some evidence of stabilization, not just a raw drop. It also considers broader market context. If the S&P is down 5% in a day and your tech stock is down 10%, that’s very different from an isolated pullback in a stable market. Another key difference is that thresholds are dynamic. A fixed “-10% from recent high” may never trigger on certain assets, or may trigger too often on others. The model adjusts relative to the asset’s own historical behavior, its range, volatility, and momentum profile -- instead of using arbitrary percentages. Finally, it doesn’t stop at “price hit X.” It suggests position sizing (e.g., scale in vs full entry) and provides an estimated entry range for limit orders based on recent volatility. So this tool is much more than price alert, but a structured execution framework: pullback + stabilization + market context + sizing + entry range. That was a good question. I'm glad you asked! Let me know if you have any more question. But the best way to learn about this tool is to sign up to test it. Thanks!