Hello Product Hunt 👋 I’m Wanis, the founder of DeskWorth. I started building DeskWorth around a simple question: can investment professionals benefit from agentic AI without giving up control over confidential data, sources and final decisions? General-purpose AI can accelerate financial research, but several problems remain: sensitive information may leave the organisation, numerical answers can be difficult to verify, sources may be unclear, and autonomous agents can create more complexity than value. DeskWorth is a local-first agentic AI environment designed for investment research. We are building it around four principles: • Private data should remain local by default. • Financial calculations should be performed by deterministic tools, not invented by a language model. • Important claims should remain connected to their original sources. • Approval and publication should remain under human control. The platform combines local models, controlled RAG, specialised financial agents, permission boundaries and traceable research workflows. External AI providers may still be useful for public information, but only through controlled and auditable routes. DeskWorth is also the applied case study behind my research on private agentic AI for investment funds. We are still testing the architecture, its limits and the operational cost of governing multi-agent systems. I would particularly value feedback from professionals working in asset management, financial research, risk, compliance, cybersecurity and AI engineering: What would you need to see before trusting an AI-generated investment analysis? Thank you for taking a look I’m looking forward to learning from the Product Hunt community.

Deskworth

Private agentic AI for verifiable investment research

Private agentic AI for verifiable investment research