The Role
Build and own Python backend systems that orchestrate multi-provider AI model pipelines for generating music videos. Improve reliability, latency, and cost, instrument production systems, contribute to prompt engineering, model evaluation, and fine-tuning, and collaborate on architecture and new product surfaces.
Summary Generated by Built In
Neural Frames is a leading AI music video generator, turning any song into a fully produced music video in minutes. We're bootstrapped, profitable, and growing fast at over $5M ARR.
Our ambition is to build a category-defining company, helping tens of millions of audio creators to visualize their content and promote it online.
As a profitable company, there is no VC drama. We are a small, ambitious group of people, striving to build something creators love.
Tasks
* Own the systems that turn songs into videos: video generation pipelines, model orchestration, queue management, and the infrastructure that keeps it all running at scale.
* Design and build backend services in Python that orchestrate multiple AI providers (we work with fal, BytePlus, OpenRouter, AWS and others) and gracefully handle their quirks, failures, and pricing trade-offs.
* Improve reliability, latency, and cost across the generation stack. Every second shaved off a render and every cent saved per video compounds across millions of generations.
* Build the infrastructure for new product surfaces working closely with the founder and tech lead on architecture decisions.
* Instrument and monitor what you ship. You'll have real visibility into how your code performs in production and the autonomy to fix what's broken.
* Contribute to the AI/ML side of the stack: prompt engineering, model evaluation, fine-tuning pipelines, and experimenting with new models as they ship.
Requirements
* You've shipped production Python backends that handled real load - async workloads, queues, retries, the unglamorous stuff that makes systems actually work.
* You're comfortable working with AI APIs and have opinions about how to wrangle them: cost, latency, failure modes, model selection.
* You think in systems. When something breaks, your instinct is to find the root cause, not patch the symptom.
* You have high agency. You see a bottleneck, you fix it. You don't wait for a ticket.
* Bonus points: experience with video/audio processing pipelines (ffmpeg, GStreamer, etc.), GPU orchestration, multi-provider AI infrastructure, or distributed systems at scale.
* You're genuinely excited about generative AI and where it's heading. You probably have a side project or two that proves it.
* You're based in Berlin and can join us in the office on at least four days a week.
Benefits
VSOPs & Profit Sharing. Bi-weekly team lunches, regular team events, an annual summer offsite, learning & development support, productivity tools, a one-time headphone budget, Deutschlandticket, and a €5,000 referral bonus.
neural frames is dedicated to creating an equitable, inclusive, and supportive work environment that brings people together from diverse backgrounds, experiences, and perspectives. Our strength is in our differences. We're proud to be an equal opportunity employer and welcome all applicants as we're committed to promoting a culture of opportunity for all.
Our ambition is to build a category-defining company, helping tens of millions of audio creators to visualize their content and promote it online.
As a profitable company, there is no VC drama. We are a small, ambitious group of people, striving to build something creators love.
Tasks
* Own the systems that turn songs into videos: video generation pipelines, model orchestration, queue management, and the infrastructure that keeps it all running at scale.
* Design and build backend services in Python that orchestrate multiple AI providers (we work with fal, BytePlus, OpenRouter, AWS and others) and gracefully handle their quirks, failures, and pricing trade-offs.
* Improve reliability, latency, and cost across the generation stack. Every second shaved off a render and every cent saved per video compounds across millions of generations.
* Build the infrastructure for new product surfaces working closely with the founder and tech lead on architecture decisions.
* Instrument and monitor what you ship. You'll have real visibility into how your code performs in production and the autonomy to fix what's broken.
* Contribute to the AI/ML side of the stack: prompt engineering, model evaluation, fine-tuning pipelines, and experimenting with new models as they ship.
Requirements
* You've shipped production Python backends that handled real load - async workloads, queues, retries, the unglamorous stuff that makes systems actually work.
* You're comfortable working with AI APIs and have opinions about how to wrangle them: cost, latency, failure modes, model selection.
* You think in systems. When something breaks, your instinct is to find the root cause, not patch the symptom.
* You have high agency. You see a bottleneck, you fix it. You don't wait for a ticket.
* Bonus points: experience with video/audio processing pipelines (ffmpeg, GStreamer, etc.), GPU orchestration, multi-provider AI infrastructure, or distributed systems at scale.
* You're genuinely excited about generative AI and where it's heading. You probably have a side project or two that proves it.
* You're based in Berlin and can join us in the office on at least four days a week.
Benefits
VSOPs & Profit Sharing. Bi-weekly team lunches, regular team events, an annual summer offsite, learning & development support, productivity tools, a one-time headphone budget, Deutschlandticket, and a €5,000 referral bonus.
neural frames is dedicated to creating an equitable, inclusive, and supportive work environment that brings people together from diverse backgrounds, experiences, and perspectives. Our strength is in our differences. We're proud to be an equal opportunity employer and welcome all applicants as we're committed to promoting a culture of opportunity for all.
Skills Required
- Shipped production Python backends handling real load, including async workloads, queues, and retries
- Design and build backend services in Python that orchestrate multiple AI providers
- Comfortable working with AI APIs and able to evaluate cost, latency, failure modes, and model selection
- Instrument, monitor, and maintain production systems, improving reliability and latency
- Contribute to AI/ML work: prompt engineering, model evaluation, and fine-tuning pipelines
- High agency; proactive problem solving and ownership of operational issues
- Based in Berlin and able to work in-office at least four days a week
- Experience with video/audio processing pipelines (ffmpeg, GStreamer)
- Experience with GPU orchestration and multi-provider AI infrastructure or distributed systems at scale
- Genuine interest in generative AI, portfolio or side projects demonstrating enthusiasm
Am I A Good Fit?
Get Personalized Job Insights.
Our AI-powered fit analysis compares your resume with a job listing so you know if your skills & experience align.
Success! Refresh the page to see how your skills align with this role.
The Company
What We Do
neural frames is a cutting-edge AI animation generator that specializes in the creation of stunning AI videos. The platform provides users with an advanced video editor and precise frame-by-frame control, offering complete creative freedom to produce high-quality animations. By combining artificial intelligence with user-friendly tools, neural frames enables creators to push the boundaries of digital art and video production.
