# Dessi Georgieva **AI Systems Engineer** London, UK · [dessi.georgieva8@gmail.com](mailto:dessi.georgieva8@gmail.com) · [LinkedIn](https://www.linkedin.com/in/dessi-georgieva/) · [GitHub](https://github.com/DG-creative-lab) · [Portfolio](https://dg-os.com/) · [ai-knowledge-hub](https://github.com/ai-knowledge-hub) --- ## Summary I architect and build enterprise AI systems that interpret user intent, coordinate skills and tools, work across hierarchical external systems, and return results under explicit identity, policy, and evidence controls. My professional work provides production backend, data, and multi-tenant experience; my independent systems make the corresponding architecture, evaluation, and recovery patterns inspectable. ## AI Systems Capabilities - **AI operating environments:** Design the people, agents, tools, data, rules, and surrounding systems that must work together to produce a useful result. - **Context, memory, and capabilities:** Give agents the right information and reusable procedures for each task, with clear sources, scope, and limits. - **Authority and human attention:** Define what an agent may decide, when it must ask, and how identity, permissions, approvals, and consequential actions stay under human and system control. - **Evaluation and learning:** Test components, interactions, and end results, then turn failures, feedback, and observed outcomes into changes that can be reviewed and reversed. - **Reliability and recovery:** Make long-running work observable and recoverable through explicit state, budgets, retries, receipts, fallback, cancellation, and rollback. ## Selected Systems ### Agentic Commerce Learning Loop _Collaborative public system · Active · [Repository](https://github.com/ai-knowledge-hub/deep-dive-analysis-agentic-commerce-augmentation)_ - I built Agentic Commerce to help brands test and improve how they describe products for paid placements and organic agent-led discovery, without giving an AI system unchecked control over business actions. - I designed a Bayesian-style learning loop that updates brand- and product-scoped beliefs as evidence arrives and carries supported patterns into later query and copy generation. It recommends whether to promote, revise, or reject each variant, while the supervised runtime keeps tools, evidence, memory, approvals, and recovery under explicit control. ### Gateplane Agent Control Plane _Independent deployed beta · Private source · [Public overview](https://gateplane-beta.vercel.app/overview)_ - I developed Gateplane independently to explore how authenticated human or agent identity becomes bounded, tenant-aware authority for tools and external effects. - The product overview is public; source code and provisioned access remain private, and employer deployment or production adoption is not claimed. ### Human Systems Platform _Founder product · Private development · [First public product](https://dg-os.com/)_ - I am developing Human Systems Platform as a founder-led product for people and organisations that need credible evidence of capability in AI-mediated work. It turns selected experience into private learning, owner-approved public evidence, and testable hypotheses about where that capability may create value. - I designed it to bring together Dessi Space for private continuity, Learning Foundry for human and agent development, DG-OS for public profiles and discovery, and a planned Organization Foundry for the context organisations need. ### Learning Foundry _Submitted public system · Judging state preserved · [Repository at submitted commit](https://github.com/DG-creative-lab/codex-hack-learning-foundry/tree/0547da02518f432fdd85e79d317e1fedaa51c4c1)_ - I built Learning Foundry as a stand-alone learning product for people who want to learn with AI without treating an agent's successful output as proof of their own understanding. It helps them explain, test, apply, and revise what they learn while separately developing evaluated agent capabilities. ## Experience ### Engineer **Performics Innovations Lab · Publicis Media** · London · Nov 2023 - Present - I map production request-to-response lifecycles and am leading architecture work for evidence-led final-answer validation; this remains in delivery rather than a deployed control. - I architect and build the Programmatic plugin and agent harness that turns ambiguous user requests into tenant-bound execution across skills, a typed CLI, backend services, analytical data, and advertising-platform APIs. - I build backend ingestion and serving workflows for heterogeneous advertising data across provider hierarchies and analytical grains, and evaluate storage architecture against query shape, aggregation, latency, reliability, and cost. - I delivered backend services and cloud data workflows for an award-recognised ecommerce optimisation platform. - I describe employer work only at responsibility and outcome level. Employer code, client information, operational measurements, and infrastructure remain confidential. ### Senior Business Intelligence Analyst **Publicis Media** · London · Mar 2023 - Nov 2023 - I bridged marketing analytics and decision systems, moving reporting workflows toward reusable intelligence services and platformised decision support. ### Business Intelligence Manager **Jellyfish** · London · Jan 2021 - Mar 2023 - I built enterprise analytics applications and data workflows, including interactive products, cloud ETL, and data-lake patterns. ### Data Consultant / SQL Developer / Data Analyst **Selected contracts** · London · 2017 - 2020 - I delivered CRM, analytics, segmentation, automation, and experimentation systems for agencies, startups, and media organisations. ## Education ### MA Applied Human Rights **University of York** · 2009 - 2011 ### BA Philosophy, specialising in Philosophy of Science **Sofia University** · 2003 - 2007