Job Title: Software Engineer III
Duration: 12 Months
Location: Remote-US (EST preferred)
Summary:

• Client is seeking a strong AI Engineer to join our Fundamental AI Research (FAIR) team, an organization focused on making research breakthroughs in AI.  Join the core team defining the future of multimodal AI. You will build the data engines and evaluation platforms that power our frontier multimodal LLMs, transforming raw data into high-fidelity benchmarks that drive breakthroughs at scale.  You’ll ship high-fidelity reproducible breakthroughs and bridge frontier research into client's product ecosystem to impact billions. You'll collaborate within an interdisciplinary group of scientists and engineers, leveraging world-class compute, data resources, and research facilities.

Must-Have Skills:

• ML/AI fundamentals - fine-tuning (SFT/preference), prompting, model-based eval, and the failure modes of each.
• At least one year of hands-on work in the field. Multimodal dataset experience - building image/video/audio/text datasets to train or evaluate models.
• Strong Python + ML tooling - PyTorch (or equivalent), HF ecosystem, comfort running training/sampling jobs on real infra.
• Web/UI engineering - production React/TypeScript; design, vibe code and ship clean, usable interfaces for annotation and internal tools.
• Data pipelines - SQL plus reliable large-scale data movement: batching, idempotent dedupe, parallelism, retries.
• Experience in mid sized tech, preference for (Google, Apple, Meta, or other FAANG)

Nice-to-have Skills:

• MS or PhD in AI or related field.
• Research track record in MLLMs - demonstrated through publications or impactful contributions to major projects.
• Open-source contributions - experience shipping and maintaining public-facing repositories.

Years of overall experience required:

• 5-10+ years of hands-on experience in building multimodal datasets, implementing machine learning evaluations, and developing production-level data and annotation tools.

Degrees/certifications required:

• Academic background in Computer Science, Engineering, or a related quantitative discipline.

What you'll do:

• Design human and model-based datasets for training and evaluating MLLMs (image/video/audio/text data).
• Run frontier model evaluations and implement common metrics.
• Use MLLMs to bootstrap, augment, and rebalance datasets.
• Fine-tune models on labeled datasets.
• Ship annotation task UIs and data-heavy internal tools (React/TypeScript)
• Run secure ingestion pipelines from warehouse/warm storage/flat files at scale.

Interview Process:

• How many rounds of interviews: 2 rounds
• Types of Interviews: Technical/Behavioral
• Interview Duration: 1 hour each

 
 
At SPECTRAFORCE, we are committed to maintaining a workplace that ensures fair compensation and wage transparency in adherence with all applicable state and local laws. This position’s starting pay is: $ 75.00/hr.