Trading and Data Engineering
TradeSync
Crypto market intelligence, connected. Explore market structure, compare timeframes and turn continuous agent research into a clearer trading picture.
A workspace for the complete research session
TradeSync brings market data, derivatives activity, macro context and collected research into one workspace. Move from the market thesis to timeframe analysis, inspect ranked opportunities, review paper outcomes and follow the reasoning produced by agent harnesses (including Hermes).
Watch / 65 seconds
Inside TradeSync
A guided tour of market intelligence, timeframe research, opportunities and the architecture connecting them. Narrated with music; sound starts when you press play. Interface captures: September 2026.
Transcript and production credits
Chase in Tech presents TradeSync. A connected workspace for crypto market research. Start with the market thesis. See the wider picture, then follow the evidence behind each scenario. Compare funding, open interest, volume and liquidation activity, with context from multiple market data providers. Explore short and long timeframes. Inspect trend, momentum and the historical observations behind the reading. Review ranked opportunities by symbol, direction and evidence quality. Track paper decisions and their outcomes. Adapters organize the inputs. Redis carries events. PostgreSQL preserves the record. The State API connects it to the dashboard. Agent harnesses, including Hermes, synthesize research and compare evidence. Scheduled workflows connect market analysis with the wider Chase OS knowledge system. Explore the architecture and source available research preview at Chase in Tech dot com. Current operation is paper research.
Synthetic narration: Kyutai Pocket TTS, Alba preset by Alba MacKenna, CC BY 4.0. Script-generated and mixed; no endorsement implied. Original synthesized score by ChaseInTech; no third-party music samples. Transitions adapted from Kenney Interface Sounds (CC0). Typography: Geist (SIL OFL).
01 / The system
From market activity
to a connected research workspace.
TradeSync brings price action, derivatives data, protocol context and external research into one system. Each layer has a clear job: collect observations, calculate features, build explanations and make the result useful to the person reviewing it.
Market data
Hyperliquid
Candles, trades, funding, open interest and order-book depth form the primary market record.
Market data
Binance + Bybit
Received liquidation events and cross-venue reference metrics add context beyond a single venue.
Fundamentals
CoinGecko + DefiLlama
BTC, ETH and SOL spot references, alongside protocol TVL, support the wider market reading.
Fundamentals
Economic calendar + FRED
Economic releases frame event risk. FRED series and release dates are optional and require the operator’s key.
Research intake
TradingView · X · YouTube
Chart ideas and selected social/video research enter through the StrikeZone browser workflow as source-linked receipts.
Research context
ChaseOS + agent harnesses
Knowledge context and bounded research jobs connect observations to explanations, comparisons and follow-up questions.
- 01
Collect & normalize
Adapters / Python
Translate source messages into versioned records with symbols, timestamps, provenance and freshness.
- 02
Stream & retain
Redis / PostgreSQL
Move events between services and retain observations, decisions and outcomes in durable storage.
- 03
Measure & interpret
Rules / research services
Calculate features, compare timeframes and rank paper opportunities against their supporting evidence.
- 04
Explore & review
State API / React
Bring the thesis, charts, research, opportunities and operating health into an interactive workspace.
Mission Control · Markets · Timeframes · Market Canvas · Opportunities · Signal Ledger · Integration Pipeline
02 / Continuous research
Agent harnesses turn collected material into useful context.
The harness coordinates scheduled collection and research jobs. Models work with the task, tools, source material and retained context supplied by the surrounding system.
- 01
Collect
Scheduled jobs gather selected market and research sources.
- 02
Retrieve
ChaseOS context and source receipts provide the material for a bounded task.
- 03
Synthesize
Agent harnesses (including Hermes) summarize observations, compare evidence and draft research.
- 04
Review
Outputs return with provenance and status for inspection in the cockpit.
Where Hermes fits
Hermes is one connected agent harness. It can explain a timeframe reading, compare research and propose follow-up work. ChaseOS supplies governed knowledge context; StrikeZone supplies collected research receipts. TradeSync keeps the source, age and status visible alongside its calculated market features.
Scheduled research supports an ongoing workflow. The Integration Pipeline shows service availability and data freshness; a schedule alone does not establish uninterrupted uptime.
Runtime and research-preview details
The dated 19 September acceptance run recorded 7/7 core stages ready and 10/10 feed heartbeats connected or answering. The public profile supports paper research; live execution remains disabled. Agent commentary is advisory. Provider access and terms belong to each operator.
First-party code is published by ChaseInTech under FSL-1.1-MIT. Private records, credentials and provider datasets are excluded from the source release.
03 / Explore the workspace
Select any frame to enlarge
Mission Control — start with the whole picture
Read the market thesis, compare the horizon map and inspect the market pulse before moving into a specific instrument or research question.
The session overview combines the Bitcoin thesis, conditional scenarios, timeframe evidence and market pulse. Each section leads to deeper analysis, while the header keeps runtime status within reach. Capture: 19 September 2026.
Integration Pipeline — follow the data through the system
Inspect the market services, event transport, storage and optional connectors. See which part of the pipeline is available and where a source or service needs attention.
Service topology connects the ingestion, processing and presentation layers. Feed heartbeats and connector states let the operator trace a missing or delayed observation to its source. Capture: 19 September 2026.
Markets — read the forces behind price
Funding, open interest, liquidation activity, volume and liquidity share a focused market view.
Compare derivatives positioning with received Binance and Bybit liquidation context. Provider labels and collection limits distinguish the available observations. Capture: 15 September 2026.
Timeframes — connect the session to the larger trend
Explore horizons from one hour to six months, compare their readings and inspect the measured features behind a selected horizon.
The horizon strip exposes agreement and disagreement across timeframes. The selected view links trend, momentum, comparable historical windows and an optional agent reading. Capture: 16 September 2026.
Opportunities — turn analysis into a review queue
Search by instrument, compare direction and evidence quality, and review candidates before exploring their detail.
The shortlist brings candidate quality, timing and expiry together. Filters and deduplication help focus the review; paper records support later outcome analysis. Capture: 15 September 2026.
04 / Engineering the system
01 / Connecting a fragmented research process
A crypto research session spans price charts, derivatives metrics, economic releases, protocol context and commentary. TradeSync brings those inputs into a connected workspace so the operator can follow a question from the overview to the underlying observation, then retain the decision and revisit its outcome.
02 / A data pipeline with a durable record
Source adapters normalize market observations and attach timestamps and provenance. Redis Streams transports events between services. PostgreSQL retains observations, derived research, paper decisions and outcomes. The State API exposes read models for the cockpit, keeping the interface independent of how each provider formats its data.
03 / Agent harnesses and the research loop
Agent harnesses (including Hermes) coordinate tools, retrieved context and model reasoning around specific research tasks. StrikeZone collects selected external research; ChaseOS supplies knowledge context. The results can explain timeframe evidence, compare sources and suggest further investigation. Each output remains linked to the material it was given, making the reasoning inspectable.
04 / Research that can be revisited
Timeframe analysis connects current measurements to historical comparisons. Opportunity and signal records support paper review and outcome tracking. The design keeps calculated features, model commentary and operator decisions distinct, so a later review can examine what was observed, how it was interpreted and what followed.
05 / An inspectable engineering project
TradeSync is published through ChaseInTech as a source-available research preview. The architecture keeps the market workspace usable when optional research connectors are unavailable, and exposes health and freshness throughout the pipeline. Current operation is paper research. The repository documents provider setup, tests, service contracts and the remaining execution work.
Technology
- Python
- TypeScript
- React
- Rust
- Docker
- PostgreSQL
- Redis Streams
- Agent harnesses
Scope and boundaries
Research preview by ChaseInTech. Market inputs combine Hyperliquid with source-labelled cross-market, protocol and macro context. Agent harnesses provide advisory research. Current operation is paper research, with live execution disabled. First-party code is available under FSL-1.1-MIT; each operator supplies their own provider access.