bygalozs.dev
CV overview

Portfolio

AI Engineer

HubTalk

Feb 2026 - Present · Remote

Design, implement and validate production conversational AI systems for voice and chat applications, combining agent architecture, evaluation, software tooling and production integrations.

Role summary

  • Engineer production conversational AI systems for voice and chat applications across multiple client use cases.
  • Work across agent architecture, evaluation, integrations and voice infrastructure to deliver reliable AI-driven customer interactions.
  • Support the full agent lifecycle from requirements and implementation through validation, deployment and production improvement.

Responsibilities

  • Translate business requirements, scripts and operational rules into modular conversational architectures using prompts, routing logic, variables, knowledge sources and behavioral guardrails.
  • Implement agent logic, extraction rules, validation flows and reusable components for complex conversational scenarios.
  • Design structured evaluation workflows covering functional tests, edge cases, adversarial scenarios, stress testing and regression validation.
  • Analyze conversation results, failure modes and production behavior to refine agent logic, reliability and conversational quality.
  • Build internal utilities for agent testing, call execution, dynamic-variable validation, evidence capture and QA workflows.
  • Integrate conversational applications with REST APIs, webhooks and external services used by client workflows.
  • Configure and troubleshoot voice infrastructure involving SIP/VoIP, trunks, call routing, connectivity and telephony services.
  • Create technical documentation, reusable methodologies and implementation guidelines for agent development, evaluation and troubleshooting.

Achievements

  • Built and validated 10+ production-ready conversational AI agents across multiple client use cases, from initial requirements through testing and deployment.
  • Developed reusable conversational-agent evaluation methods, metrics and stress-test frameworks to standardize QA across multiple agents and use cases.
  • Built and validated 10+ production-ready conversational AI agents across multiple client use cases, from initial requirements through testing and deployment.

Technologies

PythonGithubOpenAI APIGemini APIAnthropic ClaudeREST APIs / WebhooksSIP / VoIPLiveKitFusionPBXRetell AICekura