main
md 64 lines 1.37 KB
Rendered Raw
1 # Local Development
2
3 ## Prerequisites
4
5 - Java 17
6 - Maven
7 - Python 3.12+
8 - Node.js 20+
9 - Docker Desktop
10 - k3d
11 - kubectl
12 - Helm
13
14 ## Validate the Foundation
15
16 ```powershell
17 mvn clean verify
18
19 python -m py_compile ai/research-engine/app/main.py
20 python -m py_compile ai/valuation-engine/app/main.py
21 python -m py_compile ai/ranking-engine/app/main.py
22 python -m py_compile ai/portfolio-optimizer/app/main.py
23
24 helm lint infrastructure/helm/ai-investment-platform
25 ```
26
27 ## Run DEV Platform
28
29 ```powershell
30 .\platform.ps1 up DEV
31 .\platform.ps1 status DEV
32 .\platform.ps1 down DEV
33 ```
34
35 The DEV command creates a local k3d cluster named `ai-investment-dev` and installs the Helm chart into the `ai-investment` namespace.
36
37 ## Spring Boot Local Ports
38
39 When services are run directly from an IDE or Maven without overriding `SERVER_PORT`, their local defaults are:
40
41 | Service | Port |
42 | --- | ---: |
43 | api-gateway | 8080 |
44 | auth-service | 8081 |
45 | portfolio-service | 8082 |
46 | broker-service | 8083 |
47 | company-service | 8084 |
48 | research-service | 8085 |
49 | recommendation-service | 8086 |
50 | risk-service | 8087 |
51 | notification-service | 8088 |
52
53 Container and Helm deployments can override these through `SERVER_PORT`.
54
55 ## Local LLM Option
56
57 DEV configuration supports Ollama via:
58
59 ```text
60 AIP_LLM_PROVIDER=ollama
61 AIP_OLLAMA_BASE_URL=http://ollama:11434
62 ```
63
64 No model is pulled or called in the foundation iteration.