Software Engineer, Transcription

LEO Technologies, LLC

full-time

Posted on: 9/17/2025

Origin: • 🇺🇸 United States

✨ AI Apply

💰 $140,000 - $170,000 per year

Mid-LevelSenior

AnsibleAWSCloudDockerElasticSearchElixirFFmpegKubernetesNode.jsNumpyPandasPythonReactTerraformTypeScript

About the role

Lead efforts to improve transcription quality by evaluating, testing, and fine-tuning ASR models (both commercial APIs and open-source).
Build pipelines that handle speaker identification, diarization, multi-language support, and noise-robust transcription in difficult audio environments.
Develop and maintain services that integrate multiple ASR providers, ensuring resilience and flexibility across transcription workflows.
Collaborate with platform engineers to ensure seamless ingestion and persistence of transcription outputs in data pipelines.
Use data wrangling and exploratory analysis to deeply understand transcription accuracy and error patterns. - Explore and apply audio engineering techniques (denoising, voice isolation, codecs, signal processing) to improve speech clarity.
Deploy and maintain transcription-related services with basic DevOps practices, ensuring scalability and reliability.
Participate in all stages of the development lifecycle: ideation, design, prototyping, implementation, deployment, and iteration.

This is a remote, WFH role.
Strong software engineering background in fields such as Computer Science, Software Engineering, or related disciplines.
5+ years of professional development experience, with significant focus on speech processing, NLP, or transcription systems.
Proficiency in Python and comfort with system-level programming when needed.
Experience with ASR frameworks (e.g., Whisper, Kaldi, Vosk, NVIDIA NeMo, or similar).
Familiarity with audio engineering tools (e.g., ffmpeg, Sox) and denoising/voice enhancement techniques.
Knowledge of speaker diarization, speaker recognition, and multi-language ASR challenges.
Experience with data analysis and wrangling (e.g., Pandas, NumPy, Jupyter) to evaluate model performance.
Understanding of cloud deployment and DevOps basics (e.g., Docker, Kubernetes, serverless workloads).
Comfort working in a fast-paced environment with dynamic objectives and quick iteration cycles.
Demonstrated ability to work independently, make tradeoffs, and deliver results with minimal supervision.
Bonus Points: Hands-on experience fine-tuning ASR models on domain-specific datasets.
Bonus Points: Familiarity with real-time streaming pipelines for audio ingestion and transcription.
Bonus Points: Exposure to search and retrieval systems (e.g., Elasticsearch) for indexing transcribed text.
Bonus Points: Prior experience in audio forensics or noisy-channel speech analysis.
Bonus Points: Experience with applying heuristics to improve transcription results.