Decompute is a technology company.
Decompute has raised $600K across 1 funding round.
Decompute has raised $600K in total across 1 funding round.
Decompute is a technology company building decentralized AI infrastructure, with its flagship product BlackBird, the world's first on-device AI agent creation platform. It enables users to build, train, and deploy powerful AI agents locally on personal devices—offline, privately, and without cloud dependency—solving problems of latency, cost, privacy, and accessibility in AI deployment for sectors like legal, finance, research, and media.[1][2][3][5] BlackBird, powered by LaserTune optimization, supports custom model fine-tuning (e.g., DeepSeek, Qwen, Llama), rapid agent creation in minutes, unlimited local queries, and edge-ready tasks like document analysis and image generation, driving growth through free tiers with premium features like on-device fine-tuning and up to 100MB file uploads.[3][5]
The company optimizes the full AI stack—hardware/software, compilers, scheduling, and algorithms—for LLMs and AGI, emphasizing energy efficiency, scalability, and sustainability while slashing costs via decentralized on-prem and on-device solutions.[1][4]
Decompute was founded by Hina Chauhan Dixit (CEO) and Jalaj Upadhyay (CSO), leveraging their deep expertise from Apple, Microsoft, and Samsung. Hina, with over 15 years in software engineering, AI leadership, and venture investing, led engineering at Apple (Security, iCloud, AI), invested in unicorns like MosaicML (acquired by Databricks) and Stability AI at Microsoft and Samsung Next, and holds patents in AI and data integrity.[2][3] Jalaj, formerly leading AI privacy at Apple, brings academic credentials from University of Waterloo and Rutgers plus infrastructure governance experience.[2]
The idea emerged from rethinking centralized cloud paradigms, inspired by their work on efficient, private AI at Big Tech, aiming to democratize AI ownership and make it affordable via edge computing. Early traction includes launching BlackBird 1.0, with plans for 15B+ parameter model support on laptops, and positive testimonials from venture capitalists.[3][5]
Decompute rides the decentralized/edge AI trend, shifting from cloud monopolies to local processing amid rising demands for privacy (e.g., GDPR, sensitive industries), cost control, and sustainability as AI energy use surges.[2][4] Timing is ideal post-AGI hype, with LLMs needing efficient infra for real-world scaling—market forces like cloud outages, data breaches, and $100B+ training costs favor edge solutions that cut latency for real-time apps in finance, healthcare, and defense.[1][2]
By enabling data ownership and custom models, Decompute influences the ecosystem toward democratization, empowering startups, enterprises, and individuals—lowering barriers like those faced by non-cloud users and fostering innovation in on-device AGI precursors.[4][5]
Decompute is positioned to lead on-device AI with BlackBird's momentum, next expanding to 15B+ parameter models on laptops, new templates, and domains via LaserTune—potentially capturing share in a $500B+ AI infra market shifting local.[5] Trends like regulatory privacy pushes, edge hardware advances (e.g., Apple silicon), and sustainable compute will accelerate growth, evolving Decompute from pioneer to ecosystem standard-setter. This infrastructure revolution makes powerful AI accessible, tying back to its mission: AGI without boundaries, at your fingertips.
Decompute has raised $600K in total across 1 funding round.
Decompute's investors include Banyan Ventures.
Decompute has raised $600K across 1 funding round. Most recently, it raised $600K Pre-seed in October 2024.
| Date | Round | Lead Investors | Other Investors |
|---|---|---|---|
| Oct 1, 2024 | $600K Pre-seed | Banyan Ventures |