Company Description
Mihira Visual Labs is a research-focused CGI and VFX studio at the forefront of filmmaking transformation through AI- and ML-driven technologies. Specializing in the creation of full-length films, the company equips creators with advanced tools to produce high-quality storytelling in a faster and more cost-effective manner. With innovation at its core, Mihira Visual Labs enables creative exploration, ensuring that human imagination remains the key differentiator. By combining cutting-edge tech with artistic vision, the studio is shaping the future of filmmaking.
Role Overview
We are looking for a hands-on Technical Lead – AI to own and lead all AI/ML systems for our creativeproduct platform.
This person will be responsible for the architecture, development, deployment, and technical leadership of all AI capabilities across the product, including LLMs, multi-modal models, image/video generation, agentic workflows, context systems, memory, fine-tuning, model orchestration, and production AI infrastructure.
This is not a pure research role or a people-management-only role. We need a builder-leader who can design systems, write code, ship features, mentor junior engineers, and convert product ideas into working AI products. Comfortable working in a small startup team where you will both write code and lead others.
You should be able to guide junior developers, solve ambiguous technical problems, and help build an AI product from the ground up.
Key Responsibilities
- Own the end-to-end AI architecture for the product, from research and prototyping to production deployment.
- Design and build AI workflows involving LLMs, image models, video models, multi-modal models, and agentic systems.
- Develop task-based and role-based AI agents with tool use, planning, memory, context handling, and session continuity.
- Build systems for prompt orchestration, model routing, fallback logic, evaluation, and quality control.
- Integrate both paid APIs and open-source models, selecting the right approach based on quality, cost, latency, and scalability.
- Deploy and manage open-source models using GPU infrastructure, inference servers, and model-serving frameworks.
- Build and maintain context, memory, RAG, embeddings, vector database, and project/session state systems.
- Work on fine-tuning, LoRAs, adapters, dataset preparation, and model customization where needed.
- Build AI tooling that can interface with software, APIs, MCP/tool-calling systems, workflow engines, and creative pipelines.
- Collaborate with product, design, backend, and creative teams to turn complex AI capabilities into simple user-facing workflows.
- Lead and mentor junior AI/ML developers, guide problem-solving, review architecture/code, and establish AI engineering standards.
- Ensure AI systems are reliable, scalable, observable, secure, and cost-efficient in production.
Required Skills & Experience
- Strong hands-on experience with Python, PyTorch and/or TensorFlow, Hugging Face, APIs, and backend integration.
- Practical experience building products with LLMs, agents, RAG, embeddings, prompt orchestration, and tool use.
- Understanding of context management, memory systems, session handling, model routing, and evaluation frameworks.
- Experience deploying AI systems using Docker, cloud platforms, GPU servers, queues, workers, monitoring, and model-serving tools.
- Ability to work with both commercial model APIs and open-source models.
- Understanding of fine-tuning, LoRA training, adapters, dataset preparation, and inference optimization.
- Strong engineering fundamentals in databases, APIs, async workflows, system design, logging, observability, and DevOps/MLOps.
- Ability to lead technically, mentor junior engineers, and operate independently in an early-stage startup environment.
Good to Have
- Experience with image generation, video generation, multi-modal AI, diffusion models, and creative AI workflows.
- Experience building AI copilots, multi-agent systems, workflow automation tools, or node-based AI pipelines.
- Experience with ComfyUI, Diffusers, Automatic1111, InvokeAI, vLLM, Triton, TensorRT, Replicate, Fal, or similar tools.
- Experience with creative AI products such as Runway, Kling, Pika, Luma, Higgsfield, Morphic, Midjourney, Stable Diffusion, Flux, HunyuanVideo, Wan, or similar systems.
- Experience in film, VFX, animation, gaming, design tools, creator platforms, or 3D workflows.
- Familiarity with Blender, Unreal Engine, Maya, Houdini, USD, FBX, 2D-to-3D, text-to-3D, or video-to-3D workflows.
Qualifications
- 5+ years of experience
- 2+ years of working with LLMs, generative AI, or applied AI systems
- Experience building and deploying production-grade AI systems
- Strong understanding of AI application architecture, including agents, RAG, memory, context handling, and model evaluation
