Full-Stack AI Engineer, Bazak

Full-Stack AI Engineer, Bazak

Ramat Gan (Bursa) · Hybrid: 3–4 days in the office

About Bazak: The Operating System for AI Commerce

More and more consumers use AI to buy products, and brands need to close those deals themselves instead of becoming a static catalog for LLMs. Bazak turns a brand’s website into a live AI operating system: one agent that sells in ChatGPT, Claude and on the brand’s own site, then serves the customer after the purchase. We go from first call to live in production in 72 hours, with no engineering needed on the client side.

Our agents have served more than 1M messages in production. Over 40% of sessions end in a completed outcome: a booking, a lead or a resolved request. Our customers include publicly traded companies running at high scale.

The Role

You’ll design, build and run the platform behind Bazak’s agents. That covers the agent runtime, the integrations that plug into each brand’s catalog and APIs, and the channels where customers talk to the agent. The work is mostly hands-on coding, and you’ll have real influence over architecture and engineering practices. You’ll work directly with the founders and take features from idea to production quickly and reliably.

What you’ll do

• Ship product features end to end: scoping, design, implementation, testing, deployment and iteration based on real conversation data.
• Build agent capabilities, including tool calling, memory, guardrails and human handoff, and measure them with per-brand eval suites.
• Connect brands to the AI ecosystem through MCP servers, ChatGPT apps and Claude connectors, and agentic commerce protocols like ACP and UCP.
• Build back-end services and integrations in TypeScript/Node.js that turn a brand’s existing site and APIs into live inventory, pricing and checkout.
• Build front-end experiences in React and TypeScript, including the on-site agent and our back-office platform.
• Shape how we ship: CI/CD, testing standards, observability and conversation analytics.
• Use AI coding tools every day, and help set a high bar through code reviews and documentation.

Must-have

• 4+ years of professional software development experience.
• Strong TypeScript/JavaScript, with production experience in Node.js and React.
• Hands-on experience building LLM agents or conversational AI in production: tool calling, orchestration, prompt and context design.
• Solid back-end fundamentals: API design, data modeling and performance.
• A track record of owning features end to end and delivering on your own.
• High engineering standards: automated testing, thoughtful reviews and maintainable code.

Nice-to-have

• Experience with agent frameworks such as Mastra, LangGraph or the Vercel AI SDK, and with RAG or knowledge-base pipelines.
• Experience building MCP servers or ChatGPT/Claude apps, or familiarity with agentic commerce protocols (ACP, UCP, x402).
• Experience with LLM evals, guardrails or PII handling.
• Postgres/Supabase.
• GCP (Cloud Run), Terraform or other infrastructure-as-code tools.
• Product analytics and observability tools such as PostHog.
• Some Python.
• Startup or founding-engineer experience, especially building from 0 to 1.