Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/holzerjm/civichacks-demo/llms.txt

Use this file to discover all available pages before exploring further.

The EcoHack dataset (ecohack_boston_environment.txt) is a synthetic environmental quality report for Boston, Q3 2025. It contains detailed metrics on air quality, urban heat islands, water quality, waste management, climate resilience, and environmental justice impacts.

​
Dataset overview

File: data/ecohack_boston_environment.txt
Scope: City of Boston Environmental Quality Report, Q3 2025
Size: ~20 lines, 8,100+ characters
Format: Plain text, structured report

​
Data sections

​
Air quality summary

  • Boston’s AQI averaged 42 (Good) during Q3 2025, a 7% improvement over Q3 2024
  • 12 days exceeded the “Moderate” threshold (AQI > 100), concentrated in July heat wave
  • Roxbury monitoring station: 15.3 µg/m³ PM2.5 (28% above citywide average)
  • East Boston: 22 ppb NO2 near airport corridor (citywide avg: 14 ppb)
The data explicitly shows environmental justice communities face disproportionate air quality impacts.

​
Sample queries

Here are the three built-in queries for the EcoHack track: 
python scripts/demo_step2_rag.py eco 1

​
Query 1: Which Boston neighborhoods have the worst air quality and why?

This query surfaces:
  • Roxbury’s elevated PM2.5 levels (15.3 µg/m³)
  • East Boston’s NO2 concentrations near the airport
  • Heat wave impacts in July (12 days exceeding Moderate AQI)
  • The 28% disparity between Roxbury and citywide averages

​
Query 2: What are the biggest environmental justice concerns in this data?

This query identifies:
  • 35% more poor air quality days in EJ communities
  • 2.4x higher childhood asthma rates
  • 40% less green space per capita
  • 3.1x proximity to contaminated sites
  • The $12M FY2025 funding increase

​
Query 3: How is climate change specifically threatening Boston’s coastline?

This query highlights:
  • 1.5 feet of sea level rise by 2050
  • $80 billion in at-risk property
  • Affected areas: East Boston, Seaport District, Dorchester waterfront
  • 47 miles of planned coastal adaptation (8% complete)
  • 3 of 12 resilience projects finished

​
Key metrics reference

MetricValueBenchmark
Citywide AQI42 (Good)7% improvement YoY
Days exceeding Moderate12Concentrated in July
Roxbury PM2.515.3 µg/m³28% above citywide
East Boston NO222 ppb57% above citywide (14 ppb)
NeighborhoodCanopy %Temperature impact
East Boston9%8-12°F hotter
Dorchester11%8-12°F hotter
Mattapan13%8-12°F hotter
30%+ canopy areas30%+Baseline
City goal (2035)35%Requires 120K new trees
ImpactEJ community multiplier
Poor air quality days+35%
Childhood asthma hospitalization2.4x
Green space access-40% (less per capita)
Proximity to contaminated sites3.1x
FY2025 targeted funding$12M (+60% YoY)

​
Using this dataset in the web app

When you select EcoHack in the Gradio app (Step 3), the interface displays:
  • Header: ”🌿 EcoHack — Analyze Boston environmental quality, climate resilience, and environmental justice data”
  • Example questions: All three queries above as clickable examples
  • Chat responses: AI answers grounded in the specific metrics from this dataset
Try asking follow-up questions like “Which neighborhoods need the most tree planting?” or “What’s causing the CSO increase?” to explore the data interactively.

​
Querying from code

Here’s how the RAG pipeline loads and queries this dataset: 
from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, Settings
from llama_index.llms.ollama import Ollama
from llama_index.embeddings.huggingface import HuggingFaceEmbedding

# Configure local AI
Settings.llm = Ollama(model="llama3.1", request_timeout=120.0)
Settings.embed_model = HuggingFaceEmbedding(model_name="all-MiniLM-L6-v2")

# Load the EcoHack dataset
data_file = "data/ecohack_boston_environment.txt"
documents = SimpleDirectoryReader(input_files=[data_file]).load_data()

# Build vector index
index = VectorStoreIndex.from_documents(documents)
query_engine = index.as_query_engine(streaming=True, similarity_top_k=3)

# Query
response = query_engine.query(
    "Which Boston neighborhoods have the worst air quality and why?"
)
response.print_response_stream()
See Step 2: RAG with civic data for the full implementation.

Build docs developers (and LLMs) love

Get started for freeTalk to us