We're solving real problems that we've lived through ourselves—and turning them into products developers actually want to use.
Here's the thing: AI coding assistants are powerful, but they're still blind to half of what developers need. They can't debug embedded hardware. They can't safely access credentials. We're changing that.
After months of wrestling with these problems in our own consulting work, we built solutions. Then we realized—every embedded developer using AI has this problem. Every dev team using AI assistants struggles with credential chaos. So we're building commercial products, and the market is responding.
Major Milestone: ProbeCodex enters public beta on January 15, 2026. First revenue target: January 31, 2026 (1-5 paying customers). Q1 target: 50+ beta users, 5-10 paying customers, $200-500 MRR.
Want deeper technical metrics? Access our full due diligence dashboard with code velocity, bug tracking, and detailed analytics.
View Full Due Diligence Dashboard →Requires investor account
The Problem We're Solving: Picture this—you're debugging an IoT device, and your AI assistant is completely useless because it can't see what's happening on the hardware. Every embedded developer hits this wall. We did too.
Our Solution: We built the first MCP server that connects Claude Code and Cursor directly to ARM debuggers. Now AI can inspect memory, read registers, analyze RTOS threads—everything you'd do manually, but faster.
Where We Are (v0.1.18-alpha, Beta Launch Jan 15):
The Market: 6M+ embedded developers worldwide. 500K already using AI assistants. We're targeting the 50K early adopters in robotics, IoT, automotive who need this now.
The Problem We're Solving: "Hey AI, deploy the staging server." But wait—you can't give your AI assistant your AWS credentials. So you copy-paste them in Slack, store them in docs, or waste hours hunting them down. Every dev team has this nightmare. It's insecure, it's chaotic, and we lived it.
Our Solution: Credentials that unlock based on what you're working on. Assigned to deploy? You get deploy access. Task done? Access revokes automatically. AI never sees raw credentials—just executes authorized intents. It's task-based authorization that actually makes sense.
Where We Are (v0.1.1-alpha):
The Market: 27M developers globally, 5M+ already using AI assistants. If we capture 1,000 companies with 20 devs each at €40/dev/month? That's €9.6M ARR. And we're just getting started.
Why the step-up? First revenue de-risks the business model significantly. Once we prove customers will pay, the company is worth more. We're targeting first revenue by Jan 31, 2026.
Investor advantage: Invest now (pre-first customer, lower valuation) and benefit as we hit revenue milestones in Q1 2026. Early investors get better terms.
Note: Valuations update based on milestones achieved. We're transparent about where we are and where we're going. Our milestone detection system automatically tracks progress (no manual claims).
We're so confident in our metrics, we built an MCP server so your AI can access real-time company data, financials, and due diligence information. No other startup is this transparent.
An MCP (Model Context Protocol) server that gives investors AI-powered access to real-time company metrics, financial data, and due diligence information. Your AI assistant (Claude, ChatGPT, Cursor, etc.) can query our data 24/7 using natural language.
Example: "Check ProbeCodex's MRR growth this month" → Your AI calls investor.metrics_growth('month') → Instant answer: "MRR up 20%, from $400 to $480, on track for Q1 target of $500."
For anyone exploring ProbeCodex as an investment opportunity.
For verified accredited investors conducting due diligence.
For investors who have committed capital or invested in ProbeCodex.
For investors with board seats - full governance access.
We're looking for 5-10 investors to beta test the Investor MCP (starting with Public Tier access). Help us shape the world's first investor MCP server. Early testers get lifetime priority access.
Requirements: AI assistant (Claude Code, ChatGPT, Cursor, etc.) · Interest in investing in AI dev tools
24/7 access to company data without waiting for monthly updates. Your AI can audit metrics anytime.
Radical transparency builds trust. Less time answering repetitive questions. Faster investor decision-making.
No other startup offers this level of AI-powered investor access. We're setting a new standard.
We're launching public beta January 15, 2026 (8 days away) and targeting first revenue by January 31. Raising €500K-€1M to scale from beta to paying customers. Looking for investors who get developer tools, understand B2B SaaS, and want to move fast. If that's you, let's have a conversation.
Email: lee@probecodex.comFounder: Lee Sandberg
Location: Sweden
Company: Droidtech 42 AI Labs AB
Stage: Pre-seed / Seed