About the Role

We're hiring an AI Engineer to build self-improving agents for FlixStock's Self-Learning AI Platform — an agentic system that combines Generative AI and fashion expertise to produce creative output at scale. You'll design the feedback loops and evaluations that help our agents learn from every output, and own platform modules end to end alongside senior engineers.

The question that defines this role: are the agents you build measurably better every week — and are they learning to improve themselves?

What You Will Own :
• Modules: Own complete platform modules — design, build, test, deploy and maintain in production.
• Self-improving agents: Build agents that learn from outcomes, corrections and quality signals.
• Evaluations: Design evals and test sets that measure agent quality and catch regressions early.
• Context engineering: Design context for LLMs (Gemini, GPT etc.) — retrieval, memory, tools and outputs.
• Agent tracing & debugging: Trace every step and tool call; debug quality, latency and cost at the source.

What the Work Looks Like :
• Build & ship:



Write clean, scalable Python for agents and AI pipelines using AI coding tools (Claude, Cursor).
• Feedback loops: Turn customer feedback, QC reviews and production outcomes into learning signals.
• Evaluate: Build eval datasets and compare agent versions so every improvement is proven.
• Trace & debug: Root-cause failures in multi-step agent runs and add evals so they stay fixed.
• Fashion & Quality: Work with Fashion, Creative and Quality teams to encode "good" into context and evals.

What We're Looking For :
• 1–2 years of software development experience (solid internships or shipped projects count).
• Strong problem-solving ability — breaking down ambiguous problems and finding root causes.
• Hands-on experience with LLMs and Machine Learning / Deep Learning systems.
• Experience with context engineering and building AI agents or LLM workflows.
• Ability to design, debug and scale syste

📌 AI Engineer (Gurugram)
🏢 FlixStock
📍 Gurugram