Asoba Asoba
Products

Same Telemetry. Four Precise Jobs.
Built for Engineers, Operators, and Executives.

Asoba turns fragmented, vendor-locked energy telemetry into decision-grade intelligence. Every layer below operates on the same open data foundation — from universal schemas and developer APIs to fine-tuned foundation models and governed terminal execution.

Open Standard · ODSE

Universal Telemetry Standard for Distributed Energy Assets

Solar and wind assets output data in dozens of proprietary formats. ODSE (Open Data Schema for Energy) provides one open, vendor-neutral standard. It maps solar inverters, wind generation, battery storage, and meters into a clean 65-field schema.

  • Universal normalization across 25+ inverter OEMs with pip install odse
  • Eliminates 120–160 hours of bespoke per-vendor data integration engineering
  • Free, community-governed open-source standard licensed under Apache 2.0 and CC-BY-SA
  • Native support for solar PV, battery health (SOH/warranty), and market settlement wheeling
Explore ODSE Standard → ODSE Code Tutorials
ODSE Normalization Pipeline Apache 2.0
Huawei Sungrow Enphase SMA / BESS ODSE ENGINE 65-Field Spec ✓ Conformance Zero Data Loss ODS-E Record timestamp kWh_active kVArh_react battery_soh settlement
Transforms vendor-proprietary registers into standardized, auditable timeseries records ready for analytics, billing, and automated trading.
Developer Platform · Ona SDK

Programmatic Telemetry & Data Infrastructure for Energy Teams

The Ona SDK gives developers direct programmatic access to energy telemetry. You can query history, stream live 5-minute data, and pull settlement records in Python or Node.js. It features client-side validation and cached snapshots delivered in under 100 milliseconds.

  • Native Python (pip install asoba) and TypeScript/Node.js (npm i @asobacloud/sdk) packages
  • Query historical data and stream live 5-minute telemetry with resumable cursor tokens
  • Client-side validation against 65-field ODS-E schema with 6 market conformance profiles
  • Sub-100ms partner snapshots with ETag caching for site KPIs, Energy-at-Risk (EAR), and schedules
pip install asoba npm i @asobacloud/sdk
v2.4.0
Snapshot Latency
< 100 ms
ETag conditional cache
Telemetry Stream
5-Min Live
Resumable cursors
Schema Validation
65 Fields
6 Conformance profiles
OEM Adapters
25+ Brands
Huawei, Sungrow, Enphase...
from asoba import OnaClient client = OnaClient() # Fetch sub-100ms KPI snapshot with Energy-at-Risk kpis = client.partner.get_kpi_rollup(site_id='Sibaya')
Foundation Model · Nehanda v3

27B Epistemic Intelligence Grounded in Energy Physics & Market Rules

Nehanda v3 is an open-weight 27B multimodal model built for electric power grids. It operates with strict source boundaries and refuses to guess when data is missing. This prevents false answers across equipment manuals, single-line diagrams, and utility tariffs.

  • 27B parameter open-weight model available on Hugging Face for on-premise or cloud hosting
  • Strict zero-fabrication boundary: cites verified evidence or refuses queries when data is absent
  • Multimodal comprehension of raw SCADA streams, single-line schematics, and equipment manuals
  • Primary cognitive backend powering Nehanda CLI workflows and JEPA inverter anomaly detection
Explore Nehanda v3 →
Epistemic Grounding Core 27B Weights
SCADA Logs OEM Manuals Grid Tariffs Single-Lines NEHANDA v3 27B Multimodal Source Filter Zero Fabrication ✓ Cited Answer ✕ Refusal Gate Unverified Refused
Strict evidence grounding ensures zero fabricated metrics: if a data point cannot be cited from telemetry or regulation, Nehanda explicitly refuses rather than guessing.
Data Tooling & Governance · Nehanda CLI

Build Energy Data Tools With Governed SDLC Controls

Energy software development needs strict operational controls. Nehanda CLI provides a governed terminal for power data teams. It cleans messy telemetry, enforces a six-phase workflow, and saves private SQLite audit logs on your machine.

  • Automatic ODS-E telemetry normalization across 25+ inverter OEMs
  • Deterministic 6-phase SDLC state machine to prevent unverified code changes
  • 100% owned, air-gapped SQLite transcript logging for compliance
  • Connects to Ona platform telemetry and APIs via Model Context Protocol (MCP)
Explore Nehanda CLI →
node bin/nehanda-ui.mjs
❯ /phase Current SDLC Phase: plan ❯ /test ⚙ AuditCodeIntegrity ✓ AuditCodeIntegrity: 0 lifecycle violations, mock-theater tests cleared ⚙ PythonSafetyChecker ✓ PythonSafetyChecker: Bandit scan clean (0 security vulnerabilities) ● Phase updated: [test] ➔ [verify]. All mandatory safety checkers passed. ❯ SubmitTest --target=lib/scripts/odse-transform.py

How It Works: Telemetry to Business Execution

An unbroken pipeline connecting physical generation equipment to everyday executive and operational decisions. No proprietary lock-in, no bespoke integration glue code, and zero technical friction for business teams.

STAGE 1 ODSE Standard Universal Telemetry Normalizes multi-vendor data from 25+ inverter brands (Huawei, Sungrow, Enphase) into a single 65-field format. ● 25+ OEM Adapters ● Apache 2.0 Standard ● Zero Vendor Lock-in ✓ pip install odse STAGE 2 Ona Platform & SDK Intelligence & APIs Processes 5-min telemetry, evaluates Energy-at-Risk (EAR), and delivers <100ms ETag cached snapshots. ● 5-Min Telemetry Streams ● Sub-100ms ETag Cache ● 6 Conformance Profiles ✓ pip install asoba STAGE 3 MCP Protocol Standardized Bridge Secure open protocol exposing telemetry tools, financial calculations, and tariffs directly to AI models. ● Open Protocol Standard ● Tool Discovery & Schema ● Controlled AI Context ✓ Zero API glue code STAGE 4 Nehanda CLI Business Accessibility Business teams query fleet health, revenue shortfalls, and wheeling reconciliations using plain English prompts. ● Plain-English Queries ● Automated Yield Reports ● Air-Gapped SQLite Log ✓ Zero coding required
1. Raw Inverters → ODSE

Messy OEM registers from Huawei, Sungrow, or Enphase are standardized locally into open 65-field ODS-E timeseries.

2. ODSE → Ona Engine

Ona computes physics-based losses, in-memory KPI snapshots, and market settlement records in under 100ms.

3. Ona → MCP Protocol

The Model Context Protocol (MCP) securely bridges the Ona data layer to intelligent agents without bespoke glue code.

4. MCP → Business Teams

Asset managers, executives, and operators ask plain-English questions in Nehanda CLI to get verified financial reports.

Why Customers Choose Asoba

Not another monitoring dashboard. Not another generic chatbot. A fundamentally governed architecture — and it's why teams switch, whichever track brought them here.

Dimension Asoba Traditional / General AI
Data Integration ✓ Minutes to integrate major OEMs. pip install odse. Modular access via SDK. pip install asoba ✗ Per-vendor pipelines. 120–160 hrs per OEM. Vendor lock-in.
Anomaly Detection ✓ Physics-based. 5-minute telemetry. Financial impact per inverter. ✗ Threshold alerts. Manual investigation. No financial attribution.
AI / Tool Governance ✓ Governed 6-phase SDLC state machine. Air-gapped SQLite audit log. Refuses fabrication. ✗ General LLM. Unchecked agent execution, hallucinated test passes, zero audit trails.
Reporting ✓ Automated ESG, investor, and wheeling reports from existing telemetry. ✗ Weeks of manual assembly from disconnected sources.
Cost to Switch ✓ Open standard. Switch vendors without rewriting your analytics stack. ✗ Proprietary formats. Switching = full reintegration.

Ready to Unify Your Energy Infrastructure?

Whether you need to standardize multi-vendor telemetry with ODSE, integrate clean streams via the SDK, deploy Nehanda v3, or empower your business team with Nehanda CLI, we can help.

Get Help & Stay Updated