Audio AI Engineer

Dev Technology; Reston (VA); Hybrid; USD 80,000 - 160,000; Full time

Benefits offered by this job

Generous time-off policy

Remote work options for eligible proj.

Tuition reimbursement

401K with matching

Job summary

Dev Technology is seeking an applied ML engineer to own the speech-to-text model for a mobile translation capability supporting federal missions. You will adapt, compress, and optimize models for iPhone deployment, handling code-switching and transliteration decisions with careful evaluation.

The role spans data pipelines, model adaptation, deployment packaging, and rigorous testing across languages, with clear communication and documentation for stakeholders.

Qualifications

• Bachelor's degree in Computer Science, Data Science, ML, Computational Linguistics, or related field.
• Strong data-engineering background building production pipelines for large, messy audio/text data.
• Hands-on experience fine-tuning/adapting speech/audio models using LoRA/QLoRA, adapters, quantization, distillation, pruning for constrained hardware.
• Practical experience with ASR/multilanguage model development and evaluation across languages, including error analysis in real-world conditions.
• Strong Python and SQL skills; experience with PyTorch, Hugging Face Transformers/PEFT, torchaudio, librosa, or comparable tooling.
• Experience deploying and monitoring production ML systems with secure handling of sensitive data in regulated environments.
• Ability to clearly explain model behavior, tradeoffs, and limitations to technical and non-technical stakeholders.

Responsibilities

• Ingest, clean, segment, label multilingual audio and transcripts with attention to code-switching.
• Fine-tune and compress large ASR models to fit on-device constraints while preserving accuracy.
• Design language-specific packaging so weights download on demand based on use-case.
• Address loanwords and transliteration by evaluating transcription decisions.
• Build reproducible evaluation pipelines (WER/CER, latency, robustness) and report results against targets.
• Create model cards, dataset docs, and evaluation write-ups for stakeholders.

Skills

Python

SQL

ASR

Model adaptation

Data engineering

Communication

Education

Bachelor's degree in CS/DS/ML/Computational Linguistics

Tools

PyTorch

Hugging Face Transformers/PEFT

torchaudio

librosa

Job description

A Role with Purpose and Impact

This role builds the speech recognition core of a mobile translation capability supporting a government agency's national security mission. The engineer will take large, high-quality speech-to-text models spanning many language families and adapt, compress, and optimize them so they run performantly on an iPhone — including handling the reality that speakers frequently mix in borrowed English terms mid-utterance, and the model needs to make a sound call on whether to transcribe those terms in English or in the source language's own transliteration.

This is an applied ML role, not a research-only position. The strongest candidate can move fluidly from raw audio data, to model adaptation and compression experiments, to a rigorous evaluation framework — and can clearly explain what they're building, why it's better than the status quo, and how they'll know it worked.

What This Role Is (and Isn't)

This position owns the speech-to-text model — its data, its training/adaptation, its size and latency on-device, and its accuracy across languages. It does not own iOS application development, translation (source-language-to-target-language), or the Swift/AVFoundation integration layer; those are handled by a separate mobile engineering function this role will collaborate closely with.

Key Responsibilities

• Data pipelines: Ingest, clean, segment, label, and version multilingual audio and transcript data, with attention to code-switching and borrowed-word phenomena across the target language set.
• Model adaptation: Fine-tune and compress large ASR models (using LoRA/QLoRA, quantization, distillation, or other parameter-efficient and size-reduction techniques as appropriate) to fit iPhone-class memory, latency, and battery constraints, while preserving transcription quality.
• Dynamic, per-language deployment: Design model packaging so language-specific weights can be selected and downloaded on demand based on use‑case context (e.g., an operator interviewing a Chinese speaker pulls only the Chinese ASR weights).
• Loanword/transliteration handling: Build and evaluate model behavior for deciding when a borrowed English term should be transcribed as‑is versus rendered in the source language's transliteration or native equivalent.
• Evaluation: Build reproducible evaluation pipelines (word/character error rate, latency, robustness to accent/noise/speaking rate/code-switching) and clearly articulate results against defined success criteria for each language and deployment target.
• Documentation & communication: Produce clear model cards, dataset documentation, and evaluation write‑ups that let technical and non‑technical stakeholders understand what the model does, how it compares to alternatives, and what its risks and limitations are.

Required Qualifications

• Bachelor's degree in Computer Science, Data Science, Machine Learning, Computational Linguistics, or a closely related field.
• Strong data‑engineering background building production pipelines for large, messy, or unstructured audio/text datasets.
• Hands‑on experience fine‑tuning or adapting speech/audio models using parameter‑efficient methods (LoRA, QLoRA, adapters) and/or model compression techniques (quantization, distillation, pruning) for constrained hardware.
• Practical experience with ASR/speech‑to‑text model development and evaluation across multiple languages, including error analysis under real‑world conditions (accents, noise, code‑switching).
• Strong Python and SQL skills; experience with PyTorch, Hugging Face Transformers/PEFT, torchaudio, librosa, or comparable tooling.
• Experience deploying and monitoring production ML systems, with an understanding of secure handling of sensitive audio, transcripts, and derived data in a regulated environment.
• Ability to clearly explain model behavior, tradeoffs, and limitations to both technical and non‑technical stakeholders.

Preferred (Not Required)

• Prior exposure to mobile/on‑device ML deployment constraints (even without owning the mobile codebase directly).
• Experience with agentic or multi‑step workflow orchestration involving model outputs, retrieval, or human review.

The estimated salary range for this position is $80,000 - $160,000. This salary range is not a guarantee of compensation. The offered salary will be based on factors including relevant experience, geographic location, internal equity, and applicable contractual requirements. *Compensation may fall outside this range when appropriate.

Who We Are

Dev Technology is a growing IT company with an employee‑centric culture that works on mission‑critical projects for the federal government. We partner with our federal customers to deliver technology services and solutions, and to drive our client’s missions forward through innovation. We use Agile and DevSecOps principles to provide services including application development, biometrics and identity management, cloud and infrastructure optimization, IT and legacy modernization, and data management.

As a Washington Post Top Workplace award winner for the past THIRTEEN years in a row, the Top Workplaces USA for the past five years, and a recipient of the Companies As Responsive Employers (CARE) Award for the past six years, Dev Technology employees enjoy:

• Generous and flexible time‑off policy
• Flexible work schedules and telework options, including remote work availability for eligible projects
• Career development opportunities including a mentorship program, technical and management training through Dev University, hands‑on learning through DevLab, tuition reimbursement, and paid training opportunities
• Industry‑leading benefits including a choice of two health plans that include dental and vision, flexible spending account, commuter benefits, life insurance, and more
• 401K matching with a 5% matching contribution
• Regular team and company social events including our annual party, happy hours, fitness challenges, and more
• A focus on community engagement including company wide support activities, employer match for donations, and time off for volunteer efforts

Equal Opportunity Employer / Individuals with Disabilities / Protected Veterans

Dev Technology Group operates in the following states: AL, AR, AZ, CO, DC, FL, GA, ID, IL, IN, MD, MA, ME, MI, MN, MO, MS, NC, NJ, OH, OR, PA, SC, TN, TX, VA, WV.

SMS Terms and Privacy Notice

Dev Technology Group offers you the option to engage in SMS text conversations about your job application. By participating, you also understand that message frequency may vary depending on the status of your job application, and that message and data rates may apply. Please consult your carrier for further information on applicable rates and fees. Carriers are not liable for delayed or undelivered messages. Reply STOP to cancel and HELP for help. By opting‑in to receiving SMS text messages about your job application, you acknowledge and agree that your consent data, mobile number, and personal information will be collected and stored solely for the purpose of providing you with updates and information related to your job application. No mobile information will be shared with third parties/affiliates for marketing/promotional purposes. All the above categories exclude text messaging originator opt‑in data and consent; this information will not be shared with any third parties.