Skip to content

AF*

We are creating an AI Forward Deployed Engineer that helps organizations work together and onboard unknown systems and work with multiple systems.

Code Video Cerebral Valley

Core trick

Multiple agents map unfamiliar ERP schemas into a standard mid-layer format, iterating until ingestion validates.

Tags

productivityagents

Description

We are creating an AI Forward Deployed Engineer that helps organizations work together and onboard unknown systems and work with multiple systems. For now we are working with ERP systems to data ingestion, to onboard various and unknown system automatically using multiple agents which research and improve iteratively and then deploy the pipeline automatically.
Lendica-PlayGround/erp-data-ingestion★ 00 forksPythonCreated Apr 18, 2026 · last push Apr 20, 2026

README excerpt

# ERP Data Ingestion

This repository is for an agentic ERP and data-provider ingestion system that standardizes source data into a mid-layer format and prepares it for downstream loading.

## Current Scope

The current repository includes:
- Django app scaffolding under `apps/django_api/`
- **Mira** (Phase 2–3 agent) under [`mira/`](./mira/) — see [`mira/README.md`](./mira/README.md)
- Seed datasets under `seeds/`
- Product and phase specs under `docs/`

## Specs

The current project specs live in:
- [`docs/0001-prd.md`](./docs/0001-prd.md)
- [`docs/0002-phase1-midlayer-db-contract.md`](./docs/0002-phase1-midlayer-db-contract.md)
- [`docs/0005-phase4-5-observability-validation.md`](./docs/0005-phase4-5-observability-validation.md)
- [`docs/discussion/initial-discussion.md`](./docs/discussion/initial-discussion.md)

For now, these documents are the source of truth for product direction and phase-one scope.

Feature and design specs going forward should be added under [`docs/specs/`](./docs/specs/).

## Repository Layout

```text
mira/      Phase 2–3 Mira agent: LangGraph runtime, skills, framework, schemas, SQL, tests
apps/      Application code and Django project scaffolding
docs/      Product requirements, phase specs, and discussion notes
           - docs/sources/   Raw source-system data formats (e.g. Invoiced.com)
           - docs/specs/     Feature specs authored before implementation
seeds/     Example source data and manifests
           - seeds/generators/gsheets_i

Team