A serverless dashboard for Emporia Vue home energy monitors, showing amps (real electrical current) per circuit instead of the watts/kWh Emporia's own app defaults to.
I built this after finding Emporia's app answers "how much energy did I use" well but not "how close is this circuit to tripping its breaker" — which is the question that actually matters when you're staring at a breaker panel. Amps pulled straight from Emporia's own API, true hourly peaks instead of averages (a 20A spike averaged over an hour looks nothing like 20A), and an all-time peak per circuit that survives past Emporia's own retention window.
This is a portfolio copy of a project I run for my own house — see What this repo is below.
- A table of every circuit, its last-hour peak amps, and its all-time peak amps
- A 30-day graph of whole-home peak amps per hour
- An optional "live" toggle for an on-demand real-time reading (off by default, so it doesn't hammer Emporia's API on every page load)
Emporia's own API won't report amps for its combined whole-home channel —
only per-leg. So whole-home ("Main") amps has to be computed by summing the
two split-phase legs' per-second readings before taking the max — summing
each leg's own max instead overstates peaks that never happened
simultaneously. That distinction (and a handful of other non-obvious things
learned by testing directly against a real Emporia account — pagination,
warm-Lambda client caching, pyemvue's retry loop colliding with the Lambda
timeout) is documented inline in the code and in docs/.
No EC2, load balancer, NAT gateway, or API Gateway — just:
EventBridge (hourly) ──▶ PollerFunction ──▶ DynamoDB ◀── WebFunction ◀── HTTPS (Function URL)
(Lambda) (1 table) (Lambda)
poller— a Lambda on an hourly EventBridge schedule. Logs into Emporia, pulls the last hour's per-second amp data for every circuit, and writes the true hourly max (not an average) to DynamoDB. Also maintains a "LATEST" row per circuit (shown on page load) and an "ALL_TIME_MAX" row per circuit (never expires — the one place peaks survive past 30 days).web— a second Lambda, reachable directly over HTTPS via a Lambda Function URL (no web server or load balancer needed). Serves the dashboard page plus two JSON routes:GET /api/history— last-hour-peak + all-time-peak per circuit, plus 30 days of hourly whole-home history for the graph.GET /api/live— an on-demand real-time reading, only called while the page's "Live reading" toggle is on.
- DynamoDB (
EmporiaReadings) — a single table holds hourly history (auto-expired after 30 days via TTL), the latest reading per circuit, and the all-time-max per circuit (no TTL).
src/poller/handler.py hourly cron Lambda
src/web/handler.py HTTP-facing Lambda — page + /api/history + /api/live
src/web/page.py the entire dashboard page (HTML + CSS + JS) as one string
src/shared/emporia_client.py all Emporia API calls, including the Main-leg-summing logic
src/shared/dynamo.py all DynamoDB reads/writes
src/shared/secrets.py fetches Emporia login from AWS Secrets Manager
template.yaml AWS SAM template — every AWS resource this app uses
tests/ one test file per src/ file above
scripts/verify_amphours.py standalone diagnostic script, not part of the deployed app
docs/superpowers/specs/2026-08-20-emporia-lite-remodel-design.md has the
full design rationale, and docs/superpowers/plans/2026-08-21-emporia-lite-implementation.md
has the original implementation plan (useful if you want to see how a piece
was originally built, before later fixes touched it).
- An AWS account. This deploys real (small, billed) resources — DynamoDB is pay-per-request and both Lambdas are well within the free tier for a single household, but it isn't free hosting.
- AWS CLI
installed and configured (
aws configure) with credentials that can create Lambda, DynamoDB, IAM, EventBridge, and Secrets Manager resources. - AWS SAM CLI installed.
- Python 3.12.
- An Emporia Vue monitor already installed and set up in the Emporia app — this dashboard only reads from an existing Emporia account, it doesn't help with the physical install.
git clone https://github.com/Venither/emporia-lite-public.git
cd emporia-lite-public
pip install -r requirements-dev.txt
pytestaws secretsmanager create-secret \
--name emporia-lite/credentials \
--secret-string '{"EMPORIA_EMAIL":"you@example.com","EMPORIA_PASSWORD":"your-emporia-password"}'Note the ARN in the output — you'll pass it as a deploy parameter next.
sam build
sam deploy --guided--guided walks you through a stack name, AWS region, and the
EmporiaSecretArn parameter (paste the ARN from step 2) — and saves those
answers to samconfig.toml so future deploys are just sam deploy.
aws cloudformation describe-stacks \
--stack-name <your-stack-name> \
--query "Stacks[0].Outputs[?OutputKey=='WebFunctionUrl'].OutputValue" \
--output text(Also printed at the end of sam deploy --guided.)
The dashboard has no data until the poller runs at least once, and it's on an hourly EventBridge schedule. To see data right away instead of waiting up to an hour:
aws lambda invoke --function-name <your-stack-name>-PollerFunction-<suffix> /tmp/out.json(Find the exact function name with aws lambda list-functions --query "Functions[?contains(FunctionName, 'PollerFunction')].FunctionName".)
This is a public copy of a project I built and actively run for my own
house. I split it from the private repo before open-sourcing it because the
private version's README linked my live dashboard, which has no
authentication on its Function URL by design (see the comment in
template.yaml) — fine for a link only I have, not fine to publish. The
code here is otherwise identical: same 26 commits, same design decisions,
same test suite. Deploy it against your own Emporia account and AWS account
to run your own instance.