Private beta for systematic research teams

A research workspace for strategies that need evidence.

Synapse connects a chat notebook builder, deterministic recipe runs, debate-style research panels, global market context, and signal tracking into one workspace. It is built for turning a market idea into a runnable, inspectable research artifact.

Private beta now onboarding systematic research teams.

synapse.trading/app/notebooks/low-vol-momentum

You · 14:02

Build a low-vol momentum signal on the S&P 500. Use 12-1 momentum with volatility scaling, rebalance monthly, and report Sharpe & drawdown since 2005.

planner → outline ready · 6 cells · data: yfinance, FRED

coder ✓ cell 1/6 · load universe

coder ✓ cell 2/6 · compute 12-1 momentum

coder ✓ cell 3/6 · volatility scaling

tester ▶ running cell 4/6

Sharpe

1.42

Max DD

−14.8%

Turnover

87%

Research stack

finagent_openaifindatadata-mcpfruit-throwerlive market dataJupytermodel-agnostic LLMsLangGraphpandasNumPyFREDGDELTUS equitiesIndian equitiesfinagent_openaifindatadata-mcpfruit-throwerlive market dataJupytermodel-agnostic LLMsLangGraphpandasNumPyFREDGDELTUS equitiesIndian equities

The workflow

One workspace, several research loops.

The product is deliberately not one giant prompt box. Each workflow has a different contract: exploratory notebooks, repeatable recipes, structured debates, live data, and signal tracking.

  1. 01

    Pick a mode

    Chat for open-ended research, recipes for repeatable runs, debates for thesis formation.

  2. 02

    Run it grounded

    The orchestrator pulls findata, code search, and live market feeds — then writes the notebook, recipe, or debate transcript.

  3. 03

    Inspect everything

    Every run, metric, transcript, and signal lands in the workspace — searchable, comparable, and ready to promote.

Inside a session

The output is a working research object.

The goal is not a polished paragraph about markets. The goal is a notebook, recipe run, debate record, or signal view that can be opened, inspected, rerun, and improved.

  • Notebook artifacts. Generated notebooks include markdown, code, provenance metadata, validation events, and lineage extraction.
  • Recipe artifacts. YAML recipes compile into notebooks, execute through a Jupyter kernel, and persist metrics, folds, regimes, tags, and searches.
  • Decision artifacts. Trading-panel debates keep the analyst stages, evidence, verdict, and performance view together.
low_vol_momentum.ipynb
6/6 cells passed
[1]
## Low-vol momentum on the S&P 500
Build a long-only, volatility-scaled 12-1 momentum signal. Rebalance monthly. Evaluate 2005-present.
[2]
from findata.equity_prices import get_equity_prices
from findata.index_constituents import get_sp500_composition

universe = get_sp500_composition(asof="2005-01-01")
prices = get_equity_prices(universe.symbol, start_date="2005-01-01")
[3]
returns = prices["close"].pct_change()
mom_12_1 = prices["close"].pct_change(252).shift(21)
vol = returns.rolling(63).std()
signal = (mom_12_1 / vol).rank(axis=1, pct=True)
[4]
### Portfolio construction
[5]
weights = (signal > 0.8).astype(float)
weights = weights.div(weights.sum(axis=1), axis=0).fillna(0)
strategy_returns = (weights.shift(1) * returns).sum(axis=1)
Executed 6 cells · 4.2sno bias flags

Operations

Built for a stack you can inspect.

Synapse is in private beta, with an emphasis on readable boundaries, inspectable artifacts, and deployment choices that can harden with each research team.

Clear service boundaries

The frontend, orchestrator, data MCP, code MCP, and live-data server communicate over explicit HTTP, MCP, and WebSocket boundaries.

Model agnostic by design

Synapse is being built so teams can choose the model and provider that fit each workflow while the product contract stays stable.

Self-hostable stack

Synapse is designed for private deployments with clear service ownership, health checks, internal networking, and observability hooks.

Artifact history

Runs, searches, debates, costs, lineage, signal series, and notebook outputs are stored so research work can be inspected after generation.

FAQ

Straight answers.

What is Synapse today?+

Synapse is the web workspace around FinAgent. It gives users chat-driven notebook generation, deterministic recipe runs, debate records, live-data dashboards, notebook browsing, signal views, and artifact history.

Is Synapse model agnostic?+

Yes. Synapse is being built as a model-agnostic workspace so teams can choose the model and provider that fit each workflow while the interface, artifacts, and review loop stay consistent.

Where does the data come from?+

Research workflows use findata and data-mcp for documented wrappers across equity prices, macro, factors, crypto, volatility, files, and diagnostics. Live streams add market status, news, sentiment, corporate events, holdings, and governance context across supported markets.

Can I self-host it?+

Yes. Synapse is designed for private deployment, with the web app and backend services kept behind clear internal service boundaries.

How do I get access?+

Synapse is in private beta. Request early access and we will onboard research teams whose workflows match the current product.

What is planned next?+

The next development arc expands model choice, stronger backend auth and shared persistence, better cross-repo tests, richer signal promotion, and historical replay/backfill for live data.

Private beta is open.

Request access if you are building systematic strategies and want to test the research workflow with real notebooks, repeatable runs, and model-agnostic agents.