ML Engineer with 2 years of enterprise experience and a background in AI systems. Currently building a foundation in Classical ML and Deep Learning with a long-term trajectory toward Embodied Cognition and Autonomous Agents.

Former test automation engineer → Built a multi-agent workflow system from scratch → AIML.

Short-term: Classical ML, Deep Learning, PyTorch, Production ML pipelines
Mid-term: Reinforcement Learning, LLM Agents, Multi-Agent Systems
Long-term: Embodied AI, Digital Cognition, AI Safety

• Professional: Transitioning into ML Engineering roles
• Research: Substrate — an RL-based agent with homeostatic drives
    Exploring emergent behavior in artificial systems motivated by survival, curiosity, and adaptation.
• Infrastructure: ForgeLLM — fully local, fully open-source LLM inference engine
    Zero API calls. Zero subscriptions. Pure compute: quantization, pruning, sparse attention, custom CUDA kernels.

_{🇷🇺 Substrate — исследовательский проект по эмбодиментному познанию. Агенты с гомеостатическими драйвами (энергия, безопасность, любопытство) и наблюдение за эмерджентным поведением. Попытка создать минимальный субстрат для цифровой жизни.}

_{🇷🇺 ForgeLLM — полностью локальный open-source LLM-движок, созданный мной. Никаких API, никаких подписок, никакого облака — только чистая оптимизация: квантизация, pruning, sparse attention, кастомные CUDA-ядра. Цель: сделать ИИ бесплатным и доступным для каждого, кто имеет компьютер. Демократизация ИИ через оптимизацию.}

2026: Classical ML + Deep Learning + LeetCode
2027: Transformers + NLP/CV + RL Fundamentals
2028: Multi-Agent RL + Embodied AI → Master's in China
2030+: Research in Agent Cognition & AI Safety

In my work, AI is used minimally — only for routine code generation and boilerplate. All architecture, research direction, and creative decisions are human-made.

Open to ML Engineering roles, research collaborations, and mentorship in Agent-Based AI.