You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
$8,600 left to reach the funding goal.
Traditional Robotic Process Automation (RPA) and script-based screen automation break constantly whenever DOM trees, HTML class names, or responsive layout breakpoints change.
DripOff AI solves this fundamentally by engineering a sovereign, multimodal foundation agent that perceives computer and mobile screens semantically in real time. Rather than relying on brittle DOM hooks or cloud APIs that leak sensitive screen data, DripOff AI pairs localized Vision-Language Models (VLMs) with sub-10ms visual token grounding directly on local bare-metal silicon.
The operational production console is live and deployed at app.dripoffai.com/#/console.
Goal 1: Eliminate Local Memory Thrashing: I have built DripOff AI as a solo 23-year-old technical founder on a single 8GB RAM MacBook. Operating system telemetry reveals over 193,762,601 memory swap operations on the internal SSD. Compiling Apple MLX tensor models and processing high-DPI framebuffers causes continuous machine freezes.
Goal 2: Multi-OS Bare-Metal Execution: DripOff AI must execute deterministically across macOS, Windows, and Linux simultaneously. The goal of this grant is to acquire our Tier 1 Dual-Workstation Silicon Core to run parallel perception benchmarks across all three operating systems.
1. Zero-Swap Compilation: Complete elimination of SSD swap memory pressure (0 bytes swapped) during Apple MLX and TensorRT neural execution.
2. Cross-Platform Verification: Demonstrated 99.4% deterministic UI navigation accuracy running natively on macOS (Apple Silicon), Windows 11, and Ubuntu Linux.
100% Hardware Silicon Requisition ($8,600 USD total — Zero Cash Burn / Zero Salaries):
Apple Mac Studio M5 Max (macOS / iOS runtime): $5,350 — 64GB Unified Memory, 2TB SSD, 40-core GPU, 10GbE, dual centrifugal cooling. Dedicated silent host for MLX tensor perception with zero swap thrashing.
Ryzen 9 Host Platform PC (Windows & Linux runtime): $2,980 — AMD Ryzen 9 9900X 16C/32T, 64GB DDR5 RAM, 4TB PCIe Gen5 NVMe, dual-boot Windows 11 Pro and Ubuntu Linux LTS.
High-Speed KVM & Connectivity Spares: $270 — Hardware dual-machine USB-C/Thunderbolt KVM switch and high-bandwidth test lines for continuous 24/7 dual-workstation operation.
Suraj Ravindra Singh (Age 23) — Solo Technical Founder based in Ulwe, Navi Mumbai, India. Registered Micro Enterprise: DRIPOFF AI (Ministry of MSME: UDYAM-MH-33-0804622). I have engineered and deployed the full production console, WebSocket streaming runtime, and multi-OS coordinate grounding engine independently.
The primary risk is OS-level display driver API breaking changes across operating systems. This is mitigated by grounding actions directly against raw framebuffers and native kernel HID interfaces (macOS Quartz, Windows Win32 SendInput, Linux uinput) rather than framework-level accessibility wrappers.
$0. DripOff AI is 100% bootstrapped and self-funded by the founder.
There are no bids on this project.