AI investing assistant
Upload a 10-K and get a bull and bear brief. Paste an earnings report and find out whether they beat and what guidance did. Give it your free-cash-flow figures and get a DCF. Zeplik builds the analysis from the filings and numbers you provide, and when a live price or market cap is needed, it pulls from a real market-data feed rather than reciting a figure from a model's memory that could be months or years old, the failure that makes a general chatbot untrustworthy on markets.
Who it is for
- Retail and self-directed investors doing their own homework before they buy
- Analysts and associates who want a fast structured read on a filing or a print
- Founders and operators tracking public comps and their own market
- Anyone reviewing a portfolio for concentration and allocation
How it works
Bring the filing or the figures
Upload a 10-K, 10-Q, 8-K, S-1, or earnings release, or paste the holdings and cash-flow numbers you want analyzed. Nothing is assumed. The analysis is built only from what you provide, and you can add your thesis or the question you care about.
Ask the investing question
Say it plainly, for example "give me a bull and bear case on this 10-K" or "build a DCF from these cash flows." Zeplik routes to the right investing skill, and when you ask about a live price, it fetches a current quote instead of guessing.
Read the grounded analysis
You get a structured result with the reasoning shown and the source figures cited, framed as information and not a recommendation. Keep going in the same conversation, stress the assumptions, and change the inputs to see the output move.
What you can ask
Real asks in plain language, and the concrete deliverable you get back. Open the assistant to try your own; nothing runs until you send it.
Analyze this 10-K I am uploading and give me a bull and bear case.
A research brief built from the filing: the business and how it makes money, the growth drivers, the material risks pulled from the risk factors, and a balanced bull and bear thesis with the numbers cited.
Break down this quarterly earnings report, did they beat and what did guidance do?
An earnings read that compares reported results to the prior period, identifies the beats and misses in the release you pasted, and summarizes the guidance change and what management emphasized.
Build a DCF from these free-cash-flow figures, WACC, and terminal growth.
A discounted-cash-flow model using your inputs, with the projected flows, the discounting shown, a terminal value, and an implied per-share value, plus a sensitivity read on the key assumptions.
Review my portfolio and tell me if I am too concentrated, here are my holdings.
An allocation and concentration review across the positions you paste: weightings, sector and asset-class exposure, the single-name concentration risk, and what stands out.
Summarize this S-1, use of proceeds, dilution, and the risk factors.
A structured breakdown of the offering: what they will do with the money, the dilution mechanics, and the risk factors that actually matter, drawn from the S-1 you uploaded.
What is the live price and market cap of AAPL right now?
A current quote pulled from a real market-data feed with the value and the time it was retrieved, so the number is dated and checkable rather than recalled from training data.
What it can do
The assistant routes to the right skill for you. Each one below is a focused, ready-to-run workflow; follow a link to see exactly what it does.
Filings and earnings
Turn a long document into the few things that move the thesis.
- Filing analysis10-K, 10-Q, 8-K, and S-1 read for the material items and real risks.
- Earnings analysisBeat or miss, the drivers, and what guidance and the call signaled.
- Equity researchA grounded bull and bear brief on a company from its filings and figures.
Valuation and portfolio
Put a number on it, and check how the whole book is built.
- Valuation modelDCF and comparable-companies valuation from the inputs you provide.
- Portfolio reviewConcentration, allocation, and exposure across the holdings you paste.
Charts, trades, and market posture
Read a chart, size and plan a trade, and gauge the market backdrop, on figures you provide.
- Technical analystTrend, support and resistance, moving averages, and probability-weighted scenarios from a chart image.
- Position sizerShare count and dollar risk from your account size, risk percent, entry, and stop.
- Breakout plannerEntry, stop, target, and portfolio heat for a breakout or VCP setup you describe.
- Earnings reaction gradeA 5-factor momentum score of a post-earnings move from the gap, volume, and moving-average position.
- Scenario analyzerProbability-weighted base, bull, and bear paths with what would confirm each.
- Market convictionA top-down risk-on or defensive read from breadth, trend, and macro signals you supply.
- Backtest reviewStress-test a strategy for overfitting, look-ahead bias, and realistic slippage.
- Trade coachProcess, risk, and behavior review of a trading journal or closed-trade log you paste.
- Options analyzerPayoff at expiry, breakevens, max profit and loss, and Greeks intuition for an options structure you paste.
Why a general chatbot is unreliable on markets, and what is different here
Markets move every second; a language model is frozen at its training cutoff. Ask a general chatbot for a stock price and it will often state one from memory, confidently, and it can be months or years stale. On investing, a wrong number is not a rounding error, it is the whole decision. Zeplik grounds the data that changes.
The failure. Asked for a live price, market cap, or recent multiple, a general model recites a figure from its training data as if it were current, with no timestamp and no warning.
How Zeplik grounds it. When a live quote is needed, Zeplik pulls from a real market-data feed and returns the value with the time it was retrieved. If the feed is unavailable, it says so rather than reciting a stale price from memory.
The failure. A general model fills gaps in a filing with plausible-sounding numbers it did not actually read.
How Zeplik grounds it. Every research brief, earnings read, and valuation is built only from the filing or figures you provide, with the source numbers cited so you can trace each claim back to the document.
The failure. A general model slips from analysis into a recommendation, telling you to buy or sell.
How Zeplik grounds it. Every result is framed as information and analysis only. It is not investment, financial, or tax advice, and it is not a recommendation to buy or sell any security.
What it does not do
Knowing the boundary is part of using it responsibly. This is where a person, not the assistant, has to own the call.
- It is not investment advice. Everything here is information and analysis only, not a recommendation to buy or sell any security, and not financial or tax advice. Decisions and their consequences are yours.
- Live data is only as fresh as the fetch. A quote is current as of the moment it is retrieved and can move immediately after. Confirm prices and filings against a primary source before you act.
- A model is only as good as its inputs. A DCF is an opinion dressed as a number. Change the growth or discount rate and the value swings. Treat the output as a framework to stress, not an answer.
- It reads what you give it. Analysis of a filing covers the document you upload. It does not browse for the latest 8-K or the after-hours print unless you bring it or ask for a live quote.
This is information and analysis only -- it is not investment, financial, or tax advice and is not a recommendation to buy or sell any security. Figures are only as current as the data retrieved; verify prices and filings against a primary source before acting.
Frequently asked questions
- Is the AI investing assistant free?
- Zeplik is free to start and every new account includes credits. Investing skills run like any other Zeplik chat, so you spend normal usage credits on the model, shown before you send. A live quote uses a small, transparent amount for the market-data call.
- Does it give investment advice?
- No. It is information and analysis only. It is not investment, financial, or tax advice, and it is not a recommendation to buy or sell any security. Use it to research and stress-test your own thinking, then decide for yourself.
- Where do live prices come from?
- When a current price or market cap is needed, Zeplik fetches it from a real market-data feed and returns the value with the time it was retrieved, so the figure is dated rather than recalled from a model's memory. If the feed is unavailable it tells you instead of guessing.
- What can I upload?
- SEC filings like a 10-K, 10-Q, 8-K, or S-1, earnings releases, and your own holdings or cash-flow figures, as PDF, spreadsheet, or pasted text. The analysis is built only from what you provide.
- Can it build a DCF or a comps valuation?
- Yes. Give it the cash flows, discount rate, and terminal assumptions for a DCF, or peer multiples for a comparable-companies valuation, and it builds the model with the math shown and a sensitivity read on the key inputs.
- Can it analyze a whole 10-K?
- Yes. Upload the filing and it produces a structured brief: the business, the drivers, the material risks from the risk-factors section, and a balanced bull and bear case, with the figures cited back to the document.
More on Zeplik
- AI accounting and tax assistantReconcile accounts, build financials, and prep tax working papers from your exports.
- AI legal assistantReview contracts and NDAs with citations verified against a real court database.
- Every AI model, one chatRun analysis on Claude, GPT, Gemini and more, with live pricing.
Analyze your first filing
Upload a 10-K or paste your holdings and get a grounded, structured read. Free to start, with live prices from a real feed, not a model's memory.