Part II — Business Case Guide
Is ETP & meldra Right for You?
Who this is for, who it isn't for, real smart grid use cases, and an honest pros/cons list — written to help you decide, not to oversell it.
meldra is an AI agent for your smart meter Iceberg lakehouse: tell it what you need — verify a settlement figure, check feeder anomalies, investigate topology relationships — and it calls the right tools to do it. Underneath, your smart meter telemetry (AMI) lands once in your cloud storage as open Apache Iceberg — cryptographically verified at the ETP gateway, versioned, and governed — so teams across settlement, grid ops, and compliance work from the same trustworthy tables without vendor lock-in or per-query tax.
Who This Is a Good Fit For
meldra & ETP fit best if most of the following are true for you:
- You manage or ingest high-volume smart meter telemetry (AMI), SCADA logs, or feeder readings (millions to billions of rows).
- You need provable data provenance to substantiate settlement figures and tariff calculations to regulators without manual multi-week reconciliations.
- You want the underlying data in an open format you own (Apache Iceberg on your own S3/MinIO), not locked inside a vendor's proprietary storage.
- You want active ingestion security — Moving Target Defense (MTD) rotating routes and deception honeypots to defend endpoints against scanner reconnaissance.
- Your team mixes technical engineers (SQL/Python in Query Lab directly) and non-technical analysts who query via natural language grounded in catalog metadata.
- You're cost-sensitive and want to avoid per-query, per-seat pricing that platforms like Snowflake or Databricks charge once you scale usage to 50B+ rows.
Who This Is NOT a Good Fit For — Yet
Being upfront about this protects the relationship long-term.
- Companies looking for a direct procurement sale to a legacy utility procurement committee as a sole vendor without SI or OEM partners (see BC-002 Segment 1–4 strategy).
- Companies needing live legacy SCADA hardware protocols (DNP3, IEC 60870-5-104) on day one without HES integration — those require custom adapter modules.
- A buyer looking for a zero-setup consumer product — someone on the team still needs to configure gateway routes, nonces, and Iceberg table properties.
Real Use Cases
A. Verified Telemetry Ingestion — Smart Metering (UC-01)
ProblemReplay attacks, single-byte tampering, or unverified wire telemetry entering the data lake.
How ETP helpsGateways check monotonic nonces before ECDSA verification (<1ms replay rejection), committing verified blocks to
bronze_ami_readings with etp_verify_status = VERIFIED.Good fit ifyou ingest AMI telemetry from millions of edge devices and need wire-to-rest verification guarantees.
B. Settlement Substantiation & Regulatory Audit — Settlement (UC-02)
ProblemRegulators audit submitted billing/settlement figures; manual reconciliation across raw HES logs takes weeks and costs £50k–£200k.
How ETP helpsExecute settlement queries with captured Iceberg Snapshot IDs; verify partition Merkle roots against anchored RFC 3161 TSA checkpoints in seconds.
Good fit ifyour settlement team spends significant analyst time defending historical billing figures.
C. Moving Target Defense & Recon Containment — Cybersecurity (UC-03)
ProblemStatic ingestion URIs attract automated port scanners, DDoS, and credential stuffing. Blocking alerts attackers.
How ETP helpsIngress routes rotate continuously based on clock windows; invalid-route probes are silently proxied to Phantom Grid honeypots, returning synthetic responses while logging STIX threat intelligence.
Good fit ifyou operate public-facing ingestion endpoints exposed to scanner traffic.
D. Grid Event & Feeder Anomaly Investigation — Grid Operations (UC-04)
ProblemOutages or voltage excursions require fast correlation across thousands of meters to identify probable upstream transformer cause.
How ETP helpsQuery half-hourly voltage readings, push down partition filters, and traverse Apache AGE topology graphs (
GSP → Substation → Feeder → Meter) to identify the common upstream asset.Good fit ifgrid ops engineers need fast, spatial, and topological root-cause analysis during outages.
Pros and Cons, Stated Plainly
Pros
- Patented Ingestion MTD & Deception — active defense at the boundary rather than passive blocking.
- No storage lock-in — open Apache Iceberg format on your own S3/MinIO.
- Built-in Provenance & Governance — ETP verification status, daily Merkle tree checkpoints, and 7-year retention locks from day one.
- One platform, two ways in — SQL/Python in Query Lab for engineers, Chat for analysts, both calling identical bounded tools.
- Deterministic & Defensible — snapshot isolation means queries are reproducible months later.
Cons — say these out loud first
- Early-stage product. Expect active development.
- Merkle Checkpointing requires filing update — the filed patent spec describes a linear chain; Merkle trees scale verification but require claim updating within the 12-month priority window.
- Requires SQL-comfortable engineer on team for initial catalog and policy setup.