SKYNETLAB
SKYNETLAB is an independent research lab and deeptech studio focused on complex systems, spanning theoretical physics, cognitive infrastructure and socio-technical dynamics. It turns research into coherent practical tools, intellectual property and products.
At a Glance
- Professionals, researchers and founders
- Teams and organizations
- AI-agent and MCP developers
- Regulated organizations needing decision traceability
- +3 more
AI Tools by SKYNETLAB
(1)
Cortex by SKYNETLAB
Shared Memory for AI Agents
Discussions
No discussions yet
Be the first to start a discussion about SKYNETLAB
Latest News
SKYNETLAB files Italian patent application for Cortex's filtering semantic-memory architecture
Cortex by SKYNETLAB launches on Product Hunt as a hosted semantic-memory layer for AI assistants
Cortex connector repository ships a local stdio MCP bridge, Docker support and seven-tool documentation
Cortex passes OpenAI review and becomes available in the ChatGPT connector directory
Products & Services
Hosted, model-agnostic semantic memory infrastructure for AI assistants and agents. It stores decisions, context and sources across sessions, applies a write-time quality gate, tracks claims and contradictions, and exposes memory through MCP.
Coming-soon shared team memory: members use their own ChatGPT or Claude accounts while reading and writing a common, attributed memory with conflict visibility and administrator/member/reader roles.
Alpha scientific memory queryable by AI, with more than 78,000 statements from more than 15,000 publications, linking pharmacology, clinical medicine and genomics with source traceability.
Coming-soon legal and tax memory for statutes, administrative practice and judgments, represented as dated, linked claims with source references and confirmation/contradiction relationships.
Market Position
SKYNETLAB positions Cortex as a hosted, model-agnostic memory layer rather than a conventional transcript archive or vector store. Its differentiation is a write-time quality gate, typed claims, visible contradiction tracking, source citations, cross-assistant MCP access and EU infrastructure; the company's comparison page names Mem0, Zep/Graphiti, Letta, Supermemory, built-in ChatGPT memory, built-in Claude memory and DIY systems as alternatives. Cortex is hosted and closed-source/patent-pending, whereas several competitors offer open-source or self-hosted options.
Leadership
Founders
Filippo Pilotta
Independent multidisciplinary researcher and founder of SKYNETLAB, based in Bergamo/Osio Sotto, Italy. His work includes complex-systems research, theoretical physics (the Synchronia Mundi/Common Relativity project), cognitive infrastructure and AI memory; he is the inventor/applicant named on SKYNETLAB's 2026 Cortex patent application (ORCID 0009-0000-5002-4199).
Executive Team
Filippo Pilotta
Founder and independent researcher
Founder and sole operator of SKYNETLAB; multidisciplinary researcher in complex systems and theoretical physics, and inventor/applicant for the Cortex patent application.
Founding Story
SKYNETLAB was created as Filippo Pilotta's independent research lab for studying complex systems across physics, cognition and socio-technical dynamics. Its stated approach is to let research mature into usable tools, then into intellectual property and products; Cortex was built solo over two years without a team or external funding.
Business Model
Revenue Model
Subscription SaaS for hosted semantic memory. Retrieval is unlimited; paid usage is metered primarily by accepted new memories, while Team/Enterprise is offered on request with custom volumes and services.
Pricing Tiers
Formal plan provides 50 total memories, one project/brain, search/recall with sources and conflict/coherence tracking; the site says the currently open trial is temporarily wider at 1,000 memories while paid subscriptions remain closed.
100 new memories per month; signup-month and following-month starter bonus raises the cap to 200; multiple projects, agents/contexts and full data control.
500 new memories per month; signup-month and following-month starter bonus raises the cap to 1,000; includes Pro features, higher volume and processing priority.
Custom volumes, dedicated data isolation, white-label, SLA/direct support, engine configuration and dedicated onboarding.
Target Markets
- Professionals, researchers and founders
- Teams and organizations
- AI-agent and MCP developers
- Regulated organizations needing decision traceability
- Scientific, clinical and legal researchers
- Financial-market analysts and decision makers
- Persistent personal memory for AI assistants across conversations
- Project decisions, context, preferences and source-backed knowledge
- Cross-model continuity when switching between ChatGPT, Claude and other MCP clients
- Team knowledge with attribution, history and visible conflicts
- Regulated or high-accountability work requiring traceable decisions
- Scientific and medical evidence exploration