Job description
This role is for one of the Weekday's clients Salary range: Rs 500000 - Rs 1600000 (ie INR 5 - 16 LPA) Min Experience: 4 years Location: Pune, Mumbai, Gurugram, Bengaluru, Chennai, Hyderabad JobType: full-time Key Responsibilities Multi-Modal AI Pipeline: Develop and optimize AI processes for analyzing static images, media assets, video and motion content, and documents by leveraging third-party AI services, APIs, and multi-modal foundation models from major cloud hyper-scalers.
LLM Integration & Orchestration: Design, implement, and maintain an AI provider abstraction layer that supports all leading cloud AI hyper-scalers, including Google Gemini and Azure OpenAI, with intelligent routing based on task complexity and cost efficiency.
Prompt Engineering: Create and refine channel-specific prompts tailored for advertising and creative effectiveness analysis across platforms such as Facebook, Instagram, and YouTube; embed domain expertise into system prompts; and optimize prompts for structured JSON outputs from LLMs.
RAG Implementation: Develop and sustain retrieval-augmented generation workflows using pgvector, implement hybrid search combining full-text and vector similarity, and design embeddings to enhance media asset processing.
Custom AI Development: Improve our proprietary optimization model and develop algorithms that integrate third-party metrics into dynamic strength scores.
AI Agent Development: Build and oversee conversational AI agents capable of tool-calling, manage state for multi-turn dialogues, and create autonomous decision-making systems to support creative workflows.
AI-powered Asset Optimization: Oversee and advance our AI-driven solution for creative asset optimization, using creative performance analysis to automate enhancements and implement incremental edits to improve creative assets.
Performance Optimization: Minimize inference latency, implement smart caching for embeddings, optimize token usage across providers, and develop cost-effective strategies for selecting AI models.
Required Skills AI/Gen-AI Foundations: Minimum of 4 years’ experience working with LLMs (including Gemini models, GPT, Claude) and a deep understanding of embeddings and vector search techniques.
Multi-Modal AI: Demonstrated experience processing creative assets such as images and video using AI models, knowledge of computer vision fundamentals, expertise in multi-modal creative generation, and proficiency in prompt engineering for media understanding and generation.
Scaled AI Deployment: Proven experience deploying AI solutions in production, managing asynchronous processing, implementing retry mechanisms, and handling errors in AI APIs.
Python + AI Frameworks: Strong proficiency in Python, experience with cloud AI SDKs including Google Gemini, Vertex, Azure OpenAI, and familiarity with AI orchestration frameworks.
RAG Systems: Practical experience building retrieval-augmented generation solutions, utilizing vector databases, and implementing semantic search.
Creative AI Workflows: Hands-on expertise in building, managing, and evolving creative workflows powered by AI, including creative asset adaptation, automation, analysis, or production.
Highly Valued Experience in creative AI, creative technology, media, or advertising technology sectors.
Previous work with various creative AI tools and frameworks.
Advanced skills in prompt optimization and structured output formatting.
Experience fine-tuning foundation models.
Knowledge of AI safety practices, including content filtering, audit logging, and watermarking.
Understanding of advertising and media performance metrics such as engagement, recall, attention, and memorability.
AI Stack You’ll Work With AI Models: Proprietary AI APIs for creative analysis, vision models, multi-modal AI models, LLMs including Google Gemini, Azure OpenAI, and Anthropic Claude.
Internal AI: Proprietary optimization models, custom dynamic strength scoring algorithms, and channel-specific prompt libraries.
Infrastructure: PostgreSQL with pgvector (HNSW indexes), asynchronous processing pipelines, AI observability frameworks, and Model Context Protocol (MCP) for tool integration.
Frameworks: Proprietary Django-based Python framework for AI workflows and orchestration.
What You’ll Build An AI pipeline that analyzes creative assets for cognitive demand, engagement, memorability, and attention effectiveness.
A generative AI system that creates optimized creative variants based on recommendations.
A vector similarity engine that identifies conceptually similar creatives from millions of assets.
Multi-modal agents that reason across images, video, and text to deliver actionable insights.
A cost optimization layer that routes requests to the most appropriate AI models.
A compliance system (planned) for validating creatives against advertising standards.
Must-have Skills Generative AI, Azure OpenAI, Python Good-to-have Skills Creative AI Engineer, Generative AI Engineer, LLM Engineer
This job post has been translated by AI and may contain minor differences or errors.
