Artyom Makarov

I am an Applied AI professional focused on building production AI solutions, with experience supporting 60M+ monthly operations, contributing to a 10x-revenue enterprise engagement, reducing manual support by ~9,900 tickets per month, and cutting lead touchpoints by 22%. My background combines AI product development, ML engineering and data science across B2B SaaS, fintech and AI infrastructure. I work across LLMs, RAG, AI Agents, Agentic AI workflows, Python, LangChain, APIs, evaluation and production AI. Key achievements: - Built a Python and RayServe production inference pipeline that enabled an enterprise client to scale to 60M+ monthly operations and provided the technical foundation for a 10x-revenue enterprise engagement. - Designed a two-tier evaluation framework for a RAG-based AI agent, improving AI accuracy by 7% and reducing manual support escalations from 65% to 48%, equivalent to approximately 9,900 fewer manually escalated tickets per month. - Launched an ML-driven scoring and routing workflow that reduced lead touchpoints by 22% and freed up 30% of onboarding capacity for high-GMV accounts. - Reduced AI proof-of-concept cycles from 3 months to 1 month by building a hands-on Python MVP with Streamlit, LangChain and API integrations for automated corporate affiliation mapping. - Built an AI-powered Customer 360 workflow that increased B2B feedback analysis coverage from 20% to 100% at 0.5 FTE, with insights used to prioritize five strategic initiatives. - Reduced manual document verification effort by 32% through Python-based OCR matching, automated VAT verification and privacy-preserving data processing. My experience sits at the intersection of business problem solving, Applied AI engineering and production implementation. I combine hands-on technical execution with the ability to translate operational needs into AI workflows that can be evaluated, integrated and measured against real business outcomes.