top of page

AI - Conversational & Vision Engineer (Scene Generation Pilot)

Job Type

Remote (US Timing)

Experience

10+

Location

Remote (US Timing)

Job Description

Looking for a talented and innovative Conversational & Vision Engineer to lead a pivotal pilot project. You will be responsible for developing cutting-edge multimodal conversational AI capabilities that seamlessly blend natural language understanding with dynamic image generation and compositing. This contract position is key to establishing a reusable, agentic AI foundation for our future GenAI strategy.

This role centers on building a complete system that can take a natural language request, understand the context, and generate a corresponding, realistic 2D visual scene. You will leverage Amazon Bedrock and modern agentic frameworks to create a highly interactive and visually rich user experience.

Key Responsibilities

  • Build robust conversational AI backends using platforms like Amazon Bedrock and agentic orchestration frameworks (such as AgentCore or LangChain).

  • Design and implement end-to-end pipelines for 2D background scene generation and subsequent image compositing to achieve realistic visual output.

  • Develop scalable APIs to support real-time conversational queries and manage image generation requests.

  • Integrate conversational flows with various metadata or structured content sources to enhance context and fidelity.

  • Lead pilot testing initiatives focused on generation quality, latency, and overall user experience (UX).

  • Produce comprehensive technical documentation and provide recommendations for scaling the pilot system into a production environment.

Qualifications

  • Strong, hands-on experience with generative image models (Bedrock Titan Image, Stable Diffusion, SDXL).

  • Expertise in LLM-based conversation design and multi-agent orchestration using frameworks like AgentCore, LangChain, or Semantic Kernel.

  • Proficiency in Python and practical experience with AWS services for backend API development.

  • Clear understanding of multimodal embeddings and established image compositing workflows.

  • Proven ability to optimize pipelines to achieve the ideal balance between cost, performance, and visual fidelity.


Nice-to-Have Qualifications


  • Prior experience with AI applications in e-commerce, digital media, or creative industries.

  • Familiarity with 3D object rendering, PostgreSQL, or Supabase architectures.

bottom of page