Maksims B.

<contract> architected and troubleshot Omilia Conversational Platform for a UK Government Department; designed end-to-end MLOps & DevOps pipelines for LLM fine-tuning, vector-store indexing and model orchestration; acted as embedded Technical Account Manager within the Department's Platform Team. Skills: Kubernetes (Helm), Terraform, Azure DevOps & GitLab CI/CD, Python, Bash, Redis, Linux networking (TCP, SIP, RTP, TLS; Wireshark, sngrep), Solution Architecture and stakeholder management.

<startup> conceived and led the architecture of an AI-driven tender-management SaaS that applies RAG-like pro cess over domain documents (PDF, Office, CAD); hired and directed a full-stack, backend and DevOps team; built a multi-cloud DevOps/MLOps platform on Kubernetes with GitLab pipelines; integrated LLaMA-2, Falcon-40B, DeepSeek, GPT-4 and Claude via OpenAI, Azure AI & Anthropic APIs; designed data-ingestion workflows, vector DB, semantic search and Swagger/OpenAPI interfaces; since Nov 2024 serve as Board Member & principal design authority. Skills: langchain, Hugging Face, Python, Node.js, Next.js, React Native, Terraform, Helm, Postgres, Grafana/Loki, private-cloud design.

<contract> designed and implemented a DevOps framework for a CX-as-Code Genesys Cloud deployment on Azure; built Terraform-driven pipelines in Azure DevOps; integrated Cyara, Azure Cognitive Services (ASR/TTS) and Salesforce; delivered seven production contact-centre flows.