QuantTools

QuantTools is AIML Solutions’ quantitative engineering lane for multisource asset data, derivatives analytics, portfolio research, and strategy infrastructure.

It is a portfolio-facing proof point for Python quant engineering, LEAN research workflows, options/greeks analysis, Pydantic data validation, canonical JSON normalization, and agent-ready query surfaces.

What this repo does

• Executes LEAN backtests and strategy research workflows
• Ingests multisource market, derivatives, macro, broker, and on-chain records into normalized JSON and PostgreSQL
• Validates provider payloads with Pydantic before downstream analysis
• Maintains an asset-master approach for targeted equities, ETFs, crypto assets, crypto-linked ETFs, options, futures, and rates references
• Exposes query surfaces through GraphQL and JSON-RPC
• Provides hooks for MCP-enabled agent tooling
• Supports options/greeks analytics, portfolio calculations, paper-trading ledgers, and scenario research

Current implementation highlights

• Local validation suite with documented project metrics
• Multisource fixture normalization and Pydantic validation models
• Opt-in no-auth smoke checks for selected public endpoints
• Dry-run source matrix for planning provider usage without uncontrolled pulls
• LEAN, GraphQL, JSON-RPC, and local data-stack workspaces documented for continued hardening

Implementation status

Area	Status	Notes
LEAN research workflows	Scaffold/active	Strategy workspaces are included under lean-cli/; paid/cloud runs require explicit approval
Market data ingestion	Scaffold/active	Broker, macro, crypto, equities, options, and on-chain ingestion paths are organized under services/ingestion/ and services/blockchain/
Validation models	Active	Pydantic validation paths exist under services/validation/ and services/ingestion/
Query surfaces	Scaffold/active	JSON-RPC and GraphQL-facing plans/examples exist; production hardening remains a next step
Options/greeks	Scaffold/active	Framework notes and VIX/options research paths exist; expand with tested calculators and examples next
Asset master	Planned/active	A canonical target-universe model is being formalized for provider mappings, asset tags, and portfolio research
On-chain analytics	Scaffold/active	Bitcoin/Ethereum/EVM tracing and Solidity review tools are organized under services/blockchain/
Public dashboard	Active	GitHub Pages frontend artifacts summarize selected quant research outputs

This repository is the strongest current technical proof point in the AIML Solutions portfolio. See docs/PROJECT-METRICS.md for the current evidence snapshot and caveats.

Key documents

• docs/architecture.md
• docs/runbook.md
• docs/data-sources-and-market-hours.md
• docs/MULTISOURCE_ASSET_PIPELINE.md
• docs/AGENTIC_RESEARCH_CONTROL.md
• docs/SOURCE_SELECTION_STRATEGY.md
• docs/QUANTCONNECT_DATASET_MAP.md
• docs/PROJECT-METRICS.md
• docs/PRODUCTION_QUALITY_GATE.md
• docs/DATA_PIPELINE_SPEC.md
• docs/OPTIONS_GREEKS_PLAYBOOK.md
• docs/graphql-examples.md
• docs/ROADMAP.md

Quick start

# LEAN auth
lean login
lean whoami

# bring up infra
cd infra
cp .env.example .env
docker compose up -d

# run baseline local backtest
cd ../lean-cli
lean backtest "baseline-strategy" --no-update

Generate a dry-run source matrix without touching provider APIs:

python3 -m services.orchestration.source_matrix --purpose daily_baseline

Optional no-auth public smoke tests are disabled by default:

ENABLE_NETWORK_TESTS=1 python3 -m services.ingestion.network_smoke --json

Generate a compact public-safe smoke report artifact:

python3 -m services.ingestion.smoke_report --out examples/smoke-report.latest.json

Directory map

• lean-cli/ — LEAN projects + generated backtests
• lean/ — setup + ingestion scripts
• infra/ — compose + schema bootstrap
• services/validation/ — Pydantic models
• services/rpc/ — JSON-RPC scaffold
• services/mcp/ — MCP integration notes
• services/options-greeks/ — pricing framework notes
• services/blockchain/ — cross-lane chain analytics bridge

Contributing

See CONTRIBUTING.md.

License

MIT — see LICENSE.