AUTOMATED TRADING · IN DEVELOPMENT

Stock trading.
Driven by
sentiment.

Turn market commentary into signals.
Build toward automated execution.

We’re building an automated stock trading platform that combines KOL sentiment research with configurable signal criteria — from explainable research to strategy testing and future execution.

Explore the demo How it works

Research prototype. Automated execution on the roadmap.

LESS NOISE. MORE CONTEXT.01 / SENTIMENT

The sentiment pulse

SAMPLE DATA

NVIDIANVDA · Example research

Bullish

Sentiment score

75/ 100

More positive opinions than negative

7 days agoIllustrative trendToday

Bullish4

Neutral1

Bearish1

6 fictional opinionsLook behind the signal

OPINIONS → SENTIMENT → SIGNAL CRITERIAProduct prototype · In development

FROM STOCK RESEARCH TO AUTOMATION

KOL perspectives+Sentiment research+Signal criteria

01 — OUR APPROACH

Understand the sentiment.
Define the trading rules.

Our planned platform turns market commentary into structured sentiment, then evaluates that research against clearly defined strategy conditions.

Listen to the conversation.

Start with selected KOL opinions and market commentary from permissioned sources, licensed feeds, and user-provided text.

THE INPUT · MARKET VOICES

Understand the thesis.

Our planned Claude integration identifies the stock, stance, investment horizon, and reasoning — including mixed views and uncertainty.

THE ENGINE · LANGUAGE INTELLIGENCE

Shape the trading signal.

Our roadmap combines sentiment thresholds, source coverage, and market confirmation into configurable criteria for strategy testing and future execution.

THE OUTPUT · STRATEGY CONDITIONS

02 — EXPLORE THE PLATFORM

Go beyond the ticker.

Explore the sentiment layer of our planned trading platform. Choose a stock, compare opinions, and inspect the reasoning.

Sentiment workspace

INTERACTIVE PROTOTYPE

SEMICONDUCTORS

NVIDIANVDA

Sentiment score

75/100

Bullish

THE NARRATIVE

AI demand meets valuation debate.

Positive views centre on AI infrastructure demand. The opposing thesis questions how much growth is already reflected in the valuation.

Handwritten example summary · 6 fictional opinions

Balance of opinionEqual weight per opinion

Bullish 4 Neutral 1 Bearish 1

Sentiment measures opinions, not expected returns.How we calculate this

This demo runs on fictional, pre-written examples. It does not call Claude, collect KOL posts, or provide trading signals.

03 — RESEARCH INTO RULES

From sentiment research
to signal criteria.

We plan to translate sentiment into configurable trading conditions. A strategy should define when to act, how much to risk, and when to pause.

DEMO SCORING50 + 50 × (bullish − bearish) / total

0 is fully bearish. 50 is balanced. 100 is fully bullish. Neutral opinions count in the total. This simple prototype uses equal weights; it does not measure forecasting accuracy.

Define the signal.

Planned strategy settings combine sentiment thresholds, source coverage, commentary freshness, and market confirmation. Each condition should be visible before a signal is accepted.

Test before execution.

Backtesting and paper trading are planned to evaluate strategies across different market conditions before connecting them to live execution.

Keep automation controlled.

Future broker execution is intended to use user-defined position limits, stop-loss rules, and pause controls. Execution is a later roadmap stage.

04 — OUR AI ROADMAP

Language is complex.
Our ambition is clarity.

We plan to use the Claude API to turn financial commentary into structured sentiment, supporting theses, and uncertainty indicators. This research layer will inform configurable trading criteria, with evidence linked to the original material.

Talk to us about the product Claude integration planned · Product in development

THE PLANNED RESEARCH PIPELINE

INPUT

Permissioned KOL commentary

Text · Articles · Transcripts

PLANNED CLAUDE API LAYER

Extract. Classify. Explain.

EntityStock & ticker

StanceBullish / mixed / bearish

ContextThesis & time horizon

EvidenceSource & uncertainty

Human review and evaluation before production use.

05 — PRODUCT ROADMAP

Research first.
Automation next.

We’re building Automyflow to connect sentiment research with rules-based stock trading. KOL commentary provides the research layer; configurable criteria define how that research can inform a strategy.

Our roadmap progresses from live sentiment analysis and signal configuration to backtesting, paper trading, and broker-connected automation. Automated execution is a future feature, following strategy evaluation and user-controlled risk settings.

A product of AMF Business Solutions & Services Limited.

Current stageResearch prototype

A LITTLE MORE CONTEXT

Good questions.
Clear answers.

What is a KOL sentiment indicator?

A KOL is a key opinion leader. Our concept organises selected market commentators’ views about a stock into positive, neutral, and negative stances, with explanations. The indicator describes those opinions; it does not predict a stock’s price.

Is the product available today?

Automyflow is in development. You can explore the interactive research prototype on this site. Live sentiment analysis, configurable signal criteria, strategy testing, and the Claude API integration are planned. Broker-connected automated execution is a later roadmap stage.

How will Automyflow use Claude?

We plan to use the first-party Claude API to identify stock references, classify the stance of commentary, extract supporting reasoning and time horizons, and summarise opposing narratives. We intend to evaluate outputs against human-reviewed examples and preserve source evidence.

Does Automyflow execute trades?

Automated stock trade execution is on our future roadmap. We plan to connect sentiment-based signals and user-defined strategy criteria to broker integrations, with position limits, stop-loss rules, and pause controls. Development will progress through backtesting and paper trading before live execution. The current public demo shows the sentiment research layer.

LET’S CONNECT

Help shape the future
of sentiment-led trading.

Interested in sentiment research, trading strategies,
or our roadmap to automation?

Contact details will be available at launch.
For now, explore the research prototype.

Explore the prototype