Sovereign AI · Built for Trades

Train your own
model.
Own it forever.

BeardTrain collects everything your crew does — every question answered, every job stepped through, every document signed — and turns it into a fine-tuned AI that knows your shop. Runs on your hardware. Answers to you.

beardtrain — training feed
0 Examples
0 Pending Review
0 Approved
The pipeline

From job site to model weights.
Every step is yours.

01
Your apps collect data

Scout answers NEC questions. Field logs job steps. Hub transcribes voice. BeardSign processes documents. Every interaction is a training candidate.

02 🔍
Claude reviews flagged answers

When your crew thumbs down a response, Claude-p analyzes why it was wrong, what the correct answer is, and what the model should learn.

03
You approve before it sticks

Nothing enters the knowledge base without your sign-off. You review Claude's analysis, edit the correct answer if needed, and approve or reject.

04 🧠
Model gets smarter

Approved corrections feed into ShareGPT JSONL format. Your Qwen3 8B runs a fine-tuning pass on your Mac Mini. The model that comes out knows your trade.

Data sources

Every app in your stack
feeds the model.

Scout / BeardPowered AI
Electrical Q&A

NEC code questions, voltage drop calculations, wire sizing, GFCI requirements, panel sizing. Every chat is a training pair. Auto-logged on every response.

● Live
Field App
Job Step Sequences

Whisper-transcribed voice notes, parts lookups, step completions, inspection results. Real field procedures from real jobs, not textbook procedures.

● Live
Hub
Voice Intake & Task Approval

Voice-to-task conversion, dispatcher interactions, crew approvals. Teaches the model how real intake conversations flow in an electrical contracting shop.

● Live
BeardSign
Document Language

Certificate of completion language, work order descriptions, scope of work drafts. Trains the model on how trades documents are actually written.

● Live
Terminal
Infrastructure Operations

Command sequences, error fixes, deployment patterns. The model learns your exact VPS stack — PM2, nginx, SQLite, Node — from real sessions.

● Live
Pipeline / Scout CRM
Business Development

Deal conversations, proposal language, client interactions. Teaches the model the BD side of running an electrical contractor and energy development firm.

● Live
Why sovereignty matters

Your job data doesn't belong
in someone else's model.

When you ask ChatGPT about a job, that conversation leaves your shop. It becomes training data for OpenAI's next model. Your client names, your pricing, your scope of work — none of it is confidential anymore.

BeardTrain is built on the opposite premise. Your data stays on your server. Your model runs on your hardware. When you fine-tune Qwen3 8B on your job history, that model knows things no public AI ever will — because it was trained on information no one else has.

One-time build cost. No monthly AI bill. No data leaving the building.

ChatGPT / General AI
BeardTrain
Your job data
Leaves your shop
Stays on your server
NEC knowledge
Generic, sometimes wrong
Verified by your journeymen
Knows your clients
Never
Yes — trained on your CRM
Monthly cost
$20–$200/user/mo
$0 after hardware
Wrong answers
No correction loop
Human-gated correction
Model ownership
None
100% — it's your weights
The approval loop
01

Crew flags a bad answer

Any user can thumbs down a response in BeardPowered AI. One tap on the job site.

02

Claude reviews in the background

Claude-p analyzes the Q&A pair: what was wrong, what the correct answer is, what the model should learn. Takes 10–30 seconds. Runs silent.

03

You review the correction

The BeardTrain dashboard shows you the wrong answer, Claude's analysis, and the proposed correct answer. You can edit before approving.

04

Approved → knowledge base + training data

Approved corrections surface immediately in RAG (future answers get it right now) and queue for the next fine-tuning run (the model learns it permanently).

05

Fine-tune run on Mac Mini

When enough approved data accumulates, run a fine-tuning pass on Qwen3 8B locally. The model that comes out is smarter than the one that went in.

Nothing enters the knowledge base without your sign-off.

Claude is the reviewer, not the arbiter. It surfaces what it thinks is wrong and what the correct answer should be. You have the final say — every time.

This matters because NEC code changes, local AHJ interpretations vary, and field practice sometimes differs from what the codebook says. The people who know the difference are your journeymen — not an AI.

The result: a model that knows your trade the way your best journeyman does — because it was trained on your best journeyman's answers, verified by you before they became permanent.

The BeardPowered stack

Every app feeds the machine.

BeardPowered AI
ai.beard.energy
The primary field AI for electrical contractors. Answers NEC questions, sizes wire, calculates voltage drop, runs the job pipeline and CRM. Every response auto-logged as training data. Thumbs down triggers Claude-p review.
Scout
scout.beard.energy
The intelligence layer for the entire BeardPowered suite. Persistent project context, self-retrieving memory, streaming responses, app fabric integration. Built toward a fully local model on your own hardware — no external inference, no subscription.
Field App
field.beard.energy
Mobile job management for the crew. PIN login, job steps, parts photos, Whisper voice transcription, will-call emails to supply houses.
Hub
hub.beard.energy
Voice intake and task approval. Crew calls in, Hub transcribes and creates tasks. Dispatcher approves before anything hits the job board.
BeardSign
sign.beard.energy
AI-powered document signing. Certificates of completion, work orders, scope documents. AI drafts, Gemini Vision detects signature fields, both parties sign.
Contacts
contacts.beard.energy
Shared contact database across the entire suite. Clients, suppliers, inspectors, AHJ contacts. Mapbox address autocomplete. Shared across all apps.
Vault
vault.beard.energy
Knowledge graph and second brain. Force-directed graph of topics, decisions, and conversations. FTS5 search. The human-readable layer above training data.
BeardTrain
train.beard.energy
This platform. Ingests from all apps, runs Claude-p review, human approval gate, exports ShareGPT JSONL for fine-tuning Qwen3 8B on your hardware.
Terminal
terminal.beard.energy
VPS control plane. SSH terminal, file editor, PM2 management — all from mobile Safari. The coding platform the entire stack is built and deployed through.

Your crew's knowledge.
Yours to keep.

Every question your electricians ask, every job they log, every document they sign — it's all training data for a model that will outperform any general AI on your specific trade. BeardTrain makes that possible without a data science team.

Access Dashboard →