What this project shows
Stocks & Crypto AI Toolkit is a portfolio example of a real web application designed to structure market research and analysis workflows rather than imitate a decision-making machine. The system combines market data, AI, logs, dashboards and analysis history in one working structure.
Core value of the project
The project shows how Product Management, business analysis, application architecture, API integrations and AI work can be combined inside one system. AI acts here as a support layer for analysis, structured outputs and scenario review, not as a replacement for critical judgement.
The value is not only in the AI response itself, but in the full process: where the data comes from, how it is normalised, what assumptions were included in the prompt, what the execution cost was and which scenario can be revisited later.
What the system records
- candidates and scenarios,
- entry, wait and rejection conditions,
- risk checklists,
- technical levels stored as structured data,
- prompt, cost and response logs.
Problem
In market analysis the problem is rarely lack of data. It is usually too much data, too many disconnected sources and no repeatable workflow for comparing scenarios and revisiting assumptions.
Too much data
Prices, fundamentals, social signals, order book context and news appear faster than they can be reviewed manually.
Fragmented sources
Inputs come from multiple APIs, dashboards, checklists and text-based analysis layers.
No repeatable workflow
Without a shared structure it is difficult to move from broad research to a specific scenario review.
Decisions justified after the fact
If logs and scenario conditions are missing, it becomes hard to revisit why a decision was made and how strong it really was.
Product assumption
The goal was not to build a tool that tells users what to buy. The goal was to structure the research process, enforce more consistent inputs, show risks clearly and preserve the history of assumptions.
Main modules
The main page should act as a project overview. These modules show how the system separates broad monitoring, candidate research, AI workflow and auditability without turning the presentation into a full internal documentation set.
Crypto monitoring and signal review
This module structures watchlists, price data, market-cycle context, risk checklists and social intelligence for crypto assets. It helps move from broad market noise to a shortlist of situations worth further human review.
It covers watchlists, single-asset review, technical signals, order book context and narrative tracking.
Open the Crypto module pageStock and ETF research after candidate narrowing
This module shows the deeper workflow for selected stocks and ETFs: market data, fundamentals, technical snapshot, qualitative research, SWOT and bull/base/bear scenarios.
It is presented as a separate detailed module page for the later stage that follows broader screening and candidate selection.
Open the Stock Research Flow pageCandidate selection and scenario review
AI Investment Lab takes a niche or research segment and turns it into a structured process of screening, classification, shortlisting, checklists and deeper review. Each step uses constrained outputs and scenario-specific inputs.
In practice, it is the layer that turns a broad market idea into a set of candidates ready for further human-reviewed analysis.
Logs, costs and response control
A dedicated audit layer stores the prompt, model, tokens, cost, sources, execution status and JSON response. This makes it easier to revisit earlier assumptions, compare scenarios and understand how AI influenced the workflow.
It is an important portfolio element because it shows practical AI engineering rather than a single model call disconnected from business controls.
Workflow
Across modules, the project keeps the same order of work: data intake, normalisation, AI input bundle, JSON response and final review with an audit trail.
Data collection
Collecting market, technical, fundamental, social and supporting data from multiple sources.
Normalization and history
Unifying models, snapshots, cache and history so that analysis remains comparable over time.
AI input bundle
Building a prompt from structured data, scenario context, constraints and the required output format.
JSON schema output
Forcing the AI response into a schema that can be parsed further as a system object.
Review, decision, audit log
Saving the result, risks, costs, sources and decision history for later review.
Manual mode and API mode
The workflow can run manually through copy/paste prompts and JSON responses or through direct API execution with stored cost, token and status data. This makes the system more flexible than a single rigid execution path.
What this project shows about delivery capability
The project combines Product Management, Business Analysis, Django/Python, API integrations, AI engineering, prompt engineering and dashboard design for working with analysis history and scenario review.
Integrations and data sources
Instead of listing every integration separately, this section shows the four layers that make the project work: market data, narrative context, AI and the web application layer with dashboards and exports.
Market data
Binance, Twelve Data and CoinGecko provide prices, candles, indicators, asset mapping and fundamentals required for downstream analysis.
Social / narrative
LunarCrush, whale alerts and market-cycle metrics help the workflow capture narratives, sentiment, warning signals and broader scenario context.
AI layer
OpenAI API, JSON schema, PromptLog and cost logs create the layer that supports analysis, structured outputs and later workflow auditability.
App / frontend
The application layer uses Django, Django Ninja API, dashboards, DataTables, exports and working views designed for day-to-day use with data and analysis history.
Stocks & Crypto AI Toolkit screen gallery
Below are real screens from the application: crypto modules, stock views, portfolio-related screens, risk checklists, social data and AI Investment Lab workflow views.
Market risk checklist
The checklist highlights risk thresholds, statuses, data sources and elements that need review before a decision.
Stock watchlist with scoring
A watchlist of stocks and ETFs with filtering and scoring across business quality, valuation attractiveness, technical setup and risk. The view helps narrow the set of instruments worth deeper review.
Stock instrument detail
The detail view combines core market data, fundamentals, qualitative analysis and the AI-supported conclusion for one instrument.
Market cycle overview
This dashboard gathers market-cycle context such as sentiment, dominance, liquidity and macro indicators relevant for crypto scenarios.
Crypto watchlist
The watchlist aggregates tracked pairs, statuses, recommendations, signals and quick context for the crypto module.
Crypto signal watchlist
The screen combines multiple selection methods such as Wyckoff context, recommendations, StochRSI across timeframes, LunarCrush metrics, alerts, whale activity, moving averages and order book context.
Coin detail: Bitcoin
A detailed BTC view with market data, indicators, analysis sections and decision context.
AI Investment Lab run
A run detail view showing inputs, parameters, shortlisted candidates, decisions and the resulting analysis flow.
Bitget Auto Entry Monitor
The screen tracks possible LONG and SHORT entries with score, entry, SL, TP, risk-reward and execution blockers.
Would you like to discuss a similar tool?
This project is an example of building a practical AI-enabled system: from process design, data and integrations to dashboards, prompts, logs and decision-support workflow. A similar approach can support document handling, sales, customer service, reporting, CRM, ERP or internal decision dashboards.
Get in touch info@elistar.plProject type and related services
This is an internal project example showing AI and data analysis in the context of market information. It is not presented as a standard SME deployment case, but it clearly shows an approach to workflow, dashboards and AI integrations.