Job Description

Our Client is building a next-generation voice-AI platform that powers curated introductions, agentic scheduling, and real-time feedback.

We’re seeking a founding-calibre ML Engineer to own our machine learning systems end-to-end — from data pipelines and model training to evaluation and low-latency inference.

You’ll design, build, and ship ranking and recommendation systems that make every match feel more personal and improve week after week.

What You’ll Do

• Design and deploy multi-stage retrieval and re-ranking systems for compatibility, search, and personalisation.
• Build and maintain data pipelines for training, evaluation, and reporting — ensuring reproducibility and quality.
• Train and fine-tune LLMs/encoders , manage model versioning, rollout, and rollback.
• Run offline metrics (AUC, NDCG, MAP) and online A/B tests to measure real-world impact.
• Build inference services that meet tight latency and cost targets, with caching and fallback strategies.
• Implement guardrails and monitoring for drift, bias, and reliability; define model SLOs and alerts.
• Collaborate across ML, platform, and product teams to turn voice and text signals into better matches .

Tech Stack

Python • PyTorch • Hugging Face • OpenAI/Anthropic APIs • pgvector / FAISS / Pinecone • Postgres

Airflow / Prefect / dbt • AWS (S3, ECS/Kubernetes, Lambda) • CI/CD (GitHub Actions)

What We’re Looking For

• 3+ years of experience in applied ML (ranking, search, or recommendations).
• Deep Python skills and experience with PyTorch or TensorFlow (Hugging Face a plus).
• Hands-on with embeddings, LLMs, and vector search (pgvector, FAISS, Pinecone, Weaviate).
• Proven ability to take models from notebook to production — APIs, CI/CD, monitoring, rollback.
• Experience building data pipelines with Airflow, Prefect, Dagster, or dbt.
• Comfortable owning latency, cost, and reliability for inference services.
• Must be authorised to work in the U.S. and onsite 5 days/week in San Francisco.

Bonus Points

• Experience with real-time or voice systems , retrieval/re-ranking stacks, or feature stores.
• Prior work at consumer AI, dating, or recommendation startups where you shipped ranking systems.

Job Tags

Similar Jobs