feat: add Docker Compose stack with remote/cpu/single-gpu/dual-gpu profiles

.dockerignore

@@ -0,0 +1,20 @@
+.git
+__pycache__
+*.pyc
+*.pyo
+staging.db
+config/user.yaml
+config/notion.yaml
+config/email.yaml
+config/tokens.yaml
+config/craigslist.yaml
+.streamlit.pid
+.streamlit.log
+aihawk/
+docs/
+tests/
+.env
+data/
+log/
+unsloth_compiled_cache/
+resume_matcher/

.env.example

@@ -0,0 +1,19 @@
+# .env.example — copy to .env
+# Auto-generated by the setup wizard, or fill in manually.
+# NEVER commit .env to git.
+
+STREAMLIT_PORT=8501
+OLLAMA_PORT=11434
+VLLM_PORT=8000
+SEARXNG_PORT=8888
+
+DOCS_DIR=~/Documents/JobSearch
+OLLAMA_MODELS_DIR=~/models/ollama
+VLLM_MODELS_DIR=~/models/vllm
+VLLM_MODEL=Ouro-1.4B
+OLLAMA_DEFAULT_MODEL=llama3.2:3b
+
+# API keys (required for remote profile)
+ANTHROPIC_API_KEY=
+OPENAI_COMPAT_URL=
+OPENAI_COMPAT_KEY=

.gitignore

@@ -18,3 +18,4 @@ log/
 unsloth_compiled_cache/
 data/survey_screenshots/*
 !data/survey_screenshots/.gitkeep
+config/user.yaml

Dockerfile

@@ -0,0 +1,24 @@
+# Dockerfile
+FROM python:3.11-slim
+
+WORKDIR /app
+
+# System deps for companyScraper (beautifulsoup4, fake-useragent, lxml) and PDF gen
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    gcc libffi-dev curl \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Bundle companyScraper (company research web scraper)
+COPY scrapers/ /app/scrapers/
+
+COPY . .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "app/app.py", \
+     "--server.port=8501", \
+     "--server.headless=true", \
+     "--server.fileWatcherType=none"]

compose.yml

@@ -0,0 +1,83 @@
+# compose.yml — Peregrine by Circuit Forge LLC
+# Profiles: remote | cpu | single-gpu | dual-gpu
+services:
+
+  app:
+    build: .
+    ports:
+      - "${STREAMLIT_PORT:-8501}:8501"
+    volumes:
+      - ./config:/app/config
+      - ./data:/app/data
+      - ${DOCS_DIR:-~/Documents/JobSearch}:/docs
+    environment:
+      - STAGING_DB=/app/data/staging.db
+      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-}
+      - OPENAI_COMPAT_URL=${OPENAI_COMPAT_URL:-}
+      - OPENAI_COMPAT_KEY=${OPENAI_COMPAT_KEY:-}
+    depends_on:
+      searxng:
+        condition: service_healthy
+    restart: unless-stopped
+
+  searxng:
+    image: searxng/searxng:latest
+    ports:
+      - "${SEARXNG_PORT:-8888}:8080"
+    volumes:
+      - ./docker/searxng:/etc/searxng:ro
+    healthcheck:
+      test: ["CMD", "wget", "-q", "--spider", "http://localhost:8080/"]
+      interval: 10s
+      timeout: 5s
+      retries: 3
+    restart: unless-stopped
+
+  ollama:
+    image: ollama/ollama:latest
+    ports:
+      - "${OLLAMA_PORT:-11434}:11434"
+    volumes:
+      - ${OLLAMA_MODELS_DIR:-~/models/ollama}:/root/.ollama
+      - ./docker/ollama/entrypoint.sh:/entrypoint.sh
+    environment:
+      - OLLAMA_MODELS=/root/.ollama
+      - DEFAULT_OLLAMA_MODEL=${OLLAMA_DEFAULT_MODEL:-llama3.2:3b}
+    entrypoint: ["/bin/bash", "/entrypoint.sh"]
+    profiles: [cpu, single-gpu, dual-gpu]
+    restart: unless-stopped
+
+  ollama-gpu:
+    extends:
+      service: ollama
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              device_ids: ["0"]
+              capabilities: [gpu]
+    profiles: [single-gpu, dual-gpu]
+
+  vllm:
+    image: vllm/vllm-openai:latest
+    ports:
+      - "${VLLM_PORT:-8000}:8000"
+    volumes:
+      - ${VLLM_MODELS_DIR:-~/models/vllm}:/models
+    command: >
+      --model /models/${VLLM_MODEL:-Ouro-1.4B}
+      --trust-remote-code
+      --max-model-len 4096
+      --gpu-memory-utilization 0.75
+      --enforce-eager
+      --max-num-seqs 8
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              device_ids: ["1"]
+              capabilities: [gpu]
+    profiles: [dual-gpu]
+    restart: unless-stopped

docker/ollama/entrypoint.sh

@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+# Start Ollama server and pull a default model if none are present
+ollama serve &
+sleep 5
+if [ -z "$(ollama list 2>/dev/null | tail -n +2)" ]; then
+    MODEL="${DEFAULT_OLLAMA_MODEL:-llama3.2:3b}"
+    echo "No models found — pulling $MODEL..."
+    ollama pull "$MODEL"
+fi
+wait

docker/searxng/settings.yml

@@ -0,0 +1,8 @@
+use_default_settings: true
+search:
+  formats:
+    - html
+    - json
+server:
+  secret_key: "change-me-in-production"
+  bind_address: "0.0.0.0:8080"

requirements.txt

@@ -0,0 +1,63 @@
+# requirements.txt — Peregrine by Circuit Forge LLC
+# Extracted from environment.yml for Docker pip installs
+# Keep in sync with environment.yml
+
+# ── Web UI ────────────────────────────────────────────────────────────────
+streamlit>=1.35
+watchdog
+reportlab>=4.0
+pandas>=2.0
+pyarrow
+streamlit-paste-button>=0.1.0
+
+# ── Job scraping ──────────────────────────────────────────────────────────
+python-jobspy>=1.1
+playwright
+selenium
+undetected-chromedriver
+webdriver-manager
+beautifulsoup4
+requests
+curl_cffi
+fake-useragent
+
+# ── LLM / AI backends ─────────────────────────────────────────────────────
+openai>=1.0
+anthropic>=0.80
+ollama
+langchain>=0.2
+langchain-openai
+langchain-anthropic
+langchain-ollama
+langchain-community
+langchain-google-genai
+google-generativeai
+tiktoken
+
+# ── Resume matching ───────────────────────────────────────────────────────
+scikit-learn>=1.3
+rapidfuzz
+lib-resume-builder-aihawk
+
+# ── Notion integration ────────────────────────────────────────────────────
+notion-client>=3.0
+
+# ── Document handling ─────────────────────────────────────────────────────
+pypdf
+pdfminer-six
+pyyaml>=6.0
+python-dotenv
+
+# ── Utilities ─────────────────────────────────────────────────────────────
+sqlalchemy
+tqdm
+loguru
+rich
+tenacity
+httpx
+
+# ── Testing ───────────────────────────────────────────────────────────────
+pytest>=9.0
+pytest-cov
+pytest-mock
+lxml