Tech Copilot — Automotive AI Chrome Extension (Manifest V3 · Live SaaS)

Chrome Web Store Live5.0 ★ Upwork — 6 Milestones ($1,875+)Gemini Pro RAGOpenAI GPT-4o Streaming$5.99/mo Live SaaSVIN DecodingLocal OCR — Tesseract.js

"Diagnose faster. Fix it right the first time." — Tech AI-Copilot is a browser-based AI assistant built for automotive technicians. It decodes VINs, surfaces TSBs, matches symptoms to causes, and streams step-by-step repair answers — all from a side panel on any page. Built from zero to a live, monetized Chrome Web Store product by Pasindu Piumal.

Executive Summary

Tech AI-Copilot works alongside your existing systems. Decode a VIN, surface relevant TSBs and service info, connect symptoms to likely causes, and move through diagnostics with confidence — all from a side panel on any page. No tab switching. No login required to get started.

Status: Live on the Chrome Web Store
Product Website: techaicopilot.com
Pricing: Free tier + $5.99/month Pro (unlimited diagnostics, cancel anytime)
Client Rating: 5.0 ★ on Upwork over 6 paid milestones ($1,875+)

Client Testimonial

"Working with Pasindu has been one of the best decisions I've made for my startup. Tech AI-Copilot has gone from an idea to a polished product because of his technical ability, attention to detail, and willingness to tackle difficult problems. He's reliable, communicates clearly, and always works with me to find the best solution rather than the easiest one."

— Founder, Tech AI-Copilot

Real Extension UI — Screenshots

Tech AI-Copilot main panel — VIN entry, Summarize, Explain Step, Next Diagnostic Step, Write Story, Suggest Labor Time

Main panel: VIN entry + 5 quick action buttons

Tech AI-Copilot alongside OEM DTC table — Targeted Snip selecting a Diagnostic Trouble Code index

Targeted Snip: selecting a DTC index from an OEM manual

Tech AI-Copilot Snip & Scroll — capturing a multi-page OEM reference document

Snip & Scroll: capturing extended OEM reference content

Tech AI-Copilot VIN Decoding modal — 17-character VIN entry

VIN Decoding: enter a 17-char VIN, attach vehicle to session

Demo Videos — Extension in Action

Capture Method Demo

Full Screen Capture — capturing an entire OEM page into the AI chat

Targeted Snip — isolating a specific DTC table or diagnostic step

Quick Action Demos

Summarize — organizes repair tasks into labor operations

Write Story — converts diagnostic findings into an RO narrative

How It Works — Capture Methods

Two building blocks: capture the reference material you're looking at, then run a quick action to turn it into an answer, a diagnostic step, or a written-up RO.

📸 Capture Methods

Method	When to Use
Full Screen Capture	Entire page is relevant — works best when all info is on a single screen
Targeted Snip Capture	Specific table, paragraph, or DTC data point — most precise option
Snip & Scroll	Extended content across multiple screens — ideal for multi-step procedures

⚡ Quick Actions (One Tap)

Action	What It Does
Summarize	Structures multiple repair tasks into organized labor operations
Write Story	Converts technician inputs into a clear, structured RO repair narrative
Next Diagnostic Step	Evaluates symptoms + DTCs to identify the most effective next action
Explain Step	Provides reasoning behind a recommended step — not just the what, the why
Suggest Labor Time	Analyzes repair task + vehicle context for a structured labor time estimate

From Install to Answer in Under a Minute

System Architecture

4 layers

1Step 1 — Install the Extension

Add to Chrome from the Web Store in secondsNo login or account requiredSide panel opens instantlyFree tier active immediately

2Step 2 — Enter a VIN or Reference Info

Decode any 17-character VINVehicle info attaches to chat sessionAll answers grounded in correct make/model/yearOr paste reference info manually

3Step 3 — Capture + Pick a Quick Action

Full Screen, Targeted Snip, or Snip & ScrollSummarize · Explain Step · Next Diagnostic StepWrite Story · Suggest Labor TimeOr type/speak a custom question

4Step 4 — Get Instant Streamed Answers

Responses stream in real time with GPT-4oFull conversation history saved locallyPick the session back up laterNo cloud storage of your data

Engineering Architecture

Architecture & Code

┌──────────────────────────────────────────────────────────┐
│                  Chrome Side Panel (MV3)                  │
│                                                          │
│  ┌─────────────────────────────────────────────────────┐ │
│  │   React UI — VIN Input, Chat, 5 Quick Actions       │ │
│  │   • Full Screen / Targeted Snip / Snip & Scroll     │ │
│  │   • Tesseract.js Local OCR — on-device text extract │ │
│  │   • Voice Input Transcription (Web Speech API)      │ │
│  │   • Streaming LLM Chat — TSB / DTC / RO Writer      │ │
│  └──────────────────────┬──────────────────────────────┘ │
└─────────────────────────┼────────────────────────────────┘
                          ▼
┌──────────────────────────────────────────────────────────┐
│          Manifest V3 Background Service Worker           │
│                                                          │
│  ┌──────────────────────┐  ┌────────────────────────────┐│
│  │ Gemini Pro RAG        │  │ ExtensionPay Paywall       ││
│  │ + OpenAI GPT-4o SSE   │  │ $5.99/mo · Stripe Billing ││
│  │ Multi-Model Pipeline  │  │ Free Tier Quota Tracker    ││
│  └──────────────────────┘  └────────────────────────────┘│
└──────────────────────────────────────────────────────────┘

1. Chrome Side Panel Architecture (Manifest V3)

Built as a Chrome Side Panel extension — docks natively alongside any OEM manual, ALLDATA, Mitchell 1, or Identifix page without causing page reflows or CSS conflicts.

2. Three Capture Modes

Full Screen: Chrome Capture API captures the full visible viewport
Targeted Snip: Rectangle selector for precise DTC tables or spec sections
Snip & Scroll: Captures extended content across scrolling — multi-page procedures, long service manuals

3. Local OCR — Tesseract.js (On-Device)

Text extraction from screenshots runs entirely in the browser via Tesseract.js. No image ever leaves the machine for OCR — privacy-first for shop environments.

4. Multi-Model AI Pipeline

Gemini Pro Vision RAG: Analyzes captured wiring diagrams, extracts wire colors and connector pin tables
OpenAI GPT-4o SSE Streaming: Streams step-by-step diagnostic workflows in real time
Voice Input: Web Speech API transcribes spoken questions live

5. ExtensionPay Freemium ($5.99/mo Pro)

Free tier with a handful of actions to try it out
One-click Stripe subscription from within the extension
Zero custom auth backend — ExtensionPay handles all billing state

6. Privacy-First Design

OCR on-device with Tesseract.js — screenshots never leave the browser
Conversations stored locally — not in the cloud
No login required to start diagnosing
Minimal permissions model compliant with Chrome Web Store Single Purpose policy

Core Technologies

Layer	Technology
Extension	Manifest V3, Chrome Side Panel API, Chrome Capture API
AI	OpenAI GPT-4o (SSE streaming), Google Gemini Pro Vision RAG
OCR	Tesseract.js (fully on-device)
Frontend	React, Tailwind CSS, Lucide Icons
Monetization	ExtensionPay, Stripe Billing ($5.99/mo)
Voice	Web Speech API (live transcription)

Engineering Metrics & Commercial Outcomes

Metric	Manual Baseline	Tech AI-Copilot	Impact
DTC Lookup Time	3–15 min per task	Sub-500ms streamed answer	95%+ latency reduction
TSB Surface Rate	Manual search, often missed	AI-driven, surfaced automatically	Near-100% recall
OCR Processing	External service / copy-paste	On-device Tesseract.js	Zero data exposure
Monetization Infrastructure	Custom auth + billing backend	Zero-server ExtensionPay	100% cost reduction

Pricing

Plan	Price	Features
Free	$0	Handful of free actions to try it out
Pro	$5.99/month	Unlimited questions, VIN decodes, screenshots, voice input. Cancel anytime.

Frequently Asked Questions

Is Tech AI-Copilot free to use?

Yes — Tech AI-Copilot is free to install, with a handful of free actions to try it out. Unlimited access is available for $5.99/month, and you can cancel anytime.

Do I need to create an account?

No. You can open the side panel and start asking questions right away — no login is required to get going.

How does VIN decoding work?

Enter a 17-character VIN in the panel and it's decoded and attached to your chat session, so every answer is grounded in the right make, model, and year.

Where does OCR processing happen?

Text extraction from screenshots runs locally in your browser using Tesseract.js, rather than relying on an external OCR service. Your screenshots never leave your machine.

What are the capture methods?

Three options: Full Screen Capture (entire visible page), Targeted Snip (rectangle selection for precise tables or specs), and Snip & Scroll (captures multi-page content across a scrolling reference). Each feeds directly into the AI chat as context.

What quick actions are available?

Five one-tap actions: Summarize (labor op organizer), Write Story (RO narrative writer), Next Diagnostic Step (identifies the best next action from DTCs and symptoms), Explain Step (explains the why behind a recommendation), and Suggest Labor Time (structured estimate based on repair context and industry standards).

Does it replace my existing shop systems?

No — it works alongside your existing systems, helping you find relevant TSBs, service info, and diagnostic steps faster instead of starting from scratch.

Are my conversations saved?

Yes — conversations are stored locally so you can revisit, continue, or delete them whenever you like from the side panel.