Video and audio transcoding, async processing, AWS media workflow optimization.
Role Overview
Lumenairi's content is video and audio — collected on imperfect devices, in imperfect lighting, with imperfect microphones, and then transformed into broadcast-quality, multi-lingual, searchable assets. This role owns the pipeline that makes that transformation happen.
You will build and operate transcoding and audio-enhancement workflows on AWS, manage queue-based processing systems for upload-to-delivery, and continuously optimize for cost and performance. You will work alongside the Full Stack Engineer (who owns infrastructure broadly) and partner with the AI Engineer (whose captioning, transcription, and translation outputs depend on your pipeline).
Key Responsibilities
- Design and build transcoding pipelines on AWS: input from creator uploads, output to adaptive-bitrate playback for the React Native app and embeddable web players.
- Build audio enhancement workflows for podcast-quality output: noise reduction, leveling, normalization. The benchmark is consistent quality regardless of source device.
- Implement queue-based async processing: SQS, SNS, EventBridge, or equivalent. Jobs must be observable, retryable, and idempotent.
- Manage the full media lifecycle: ingest, validate, transcode, enhance, store, deliver, retire. Each transition should be auditable.
- Evaluate and integrate media tooling: FFmpeg for core processing, plus specialized tools (open-source alternatives to Topaz video enhancement are being evaluated).
- Partner with the AI Engineer to pass clean assets to AWS Transcribe (or successor) for captioning and to the search/translation pipelines.
- Continuously optimize for cost: AWS spend on transcoding, storage tiering for hot vs. cold content, delivery cost via CloudFront.
- Support future roadmap including landscape and portrait orientation handling, multi-camera depth data ingestion for future holographic playback, and per-tenant processing budgets.
Required Qualifications
- 5+ years of backend engineering experience in Node.js or Python.
- Production async processing experience: job queues, worker pools, idempotency, retry/backoff patterns.
- FFmpeg or equivalent media tooling experience. You have written or maintained transcoding scripts at production scale.
- AWS media services experience: S3, Elemental MediaConvert, Elastic Transcoder, or self-managed FFmpeg on Lambda/EC2/Fargate.
- Cost-conscious engineering — you have made tradeoffs between processing quality, latency, and AWS bill.
Nice to Have
- Audio engineering background or strong audio processing exposure.
- Experience with adaptive bitrate streaming (HLS, DASH) for mobile playback.
- Captioning and transcription pipeline experience (AWS Transcribe or similar).
- Familiarity with video super-resolution or enhancement tooling (Topaz, Real-ESRGAN, or similar).
- Design sensibility — comfort thinking about end-user perceived quality (latency, audio clarity, visual artifacts) and not just pipeline correctness.
Success Signals
- Has built real media pipelines at scale — not toy projects, but systems where reliability mattered.
- Optimizes for the right metric. Knows when latency wins and when cost wins.
- Designs for observability before the first incident, not after it.
Reporting & Working Model
Reports to the CTO. Coordinates closely with the Full Stack Engineer (infrastructure) and the AI Engineer (downstream consumption).
Pay: ₹400,000.00 - ₹1,000,000.00 per year
Benefits:
- Paid sick time
- Paid time off
- Work from home
Work Location: Hybrid remote in Mohali, Punjab
