Reward Modeling

Reward Modeling Developer Jobs

2 jobs found

Today

Machine Learning Engineer - E-Commerce Recommendation Live

TikTok Seattle, WA

Seattle, WA

Today

$207k - $368k/yr Full time Data Pipeline Generative Ai Engineer Generative Recommendations

Apply Now

Job Description

Join TikTok in Seattle to build the end-to-end recommendation stack powering TikTok Shop LIVE.

Responsibilities

Scale Large Recommendation Models (LRM): Push scaling for a unified ranking foundation model, including model capacity, sparse MoE, ultra-long user sequences, and pre-training with teacher-student distillation.
Convert offline scaling gains into online wins under strict latency and compute budgets using dynamic compute allocation and on-device real-time inference.
Generative recommendation on multimodal Semantic IDs: Design semantic IDs that fuse visual, speech, text, collaborative, and product signals (including what is being showcased, pinned, or auctioned in a live room).
Power generative retrieval with near-real-time SID serving.
LLM4Rec and MLLM cross-domain modeling: Use multimodal LLMs to represent live rooms, short videos, and products in one semantic space.
Inject world knowledge to transfer user interests across content, search, and shopping, and address cold start for new creators, new products, and new markets.
Reward models and post-training for recommenders: Build reward models for conversion, long-term value, and user experience; align recommenders with SFT and DPO/GRPO-style RL post-training.
Apply generator-evaluator decoding and listwise objectives for training and optimization.
Agent x Recommendation: Build LLM agents to auto-tune multi-objective value weights and traffic mechanisms, design and read experiments, and diagnose online cases.
Move from hand-tuned heuristics toward agentic optimization and multi-step RL.
Long-term value and ecosystem: Model LTV, uplift, and multi-touch attribution; design value and traffic mechanisms for live auctions, brands, and emerging creators.
Balance short-term GMV with user experience and a healthy, growing creator ecosystem.
Own it end to end: Drive work from problem framing and data pipelines through large-scale training, online serving, A/B testing, and launch; partner with engineering, product, and data science.

Requirements

Education: Bachelor’s degree or above in Computer Science, Electrical Engineering, Mathematics, Statistics, or a related field.
Experience: 3+ years in recommendation, search, advertising, or other large-scale applied machine learning with a record of shipped models that improved online metrics.
ML depth: Strong machine learning fundamentals and hands-on experience in at least one area such as retrieval and ranking, sequential user modeling, multi-task learning, generative recommendation, or LLM/multimodal modeling.
Programming + frameworks: Proficiency in Python and/or C++, plus solid experience with PyTorch or TensorFlow, including distributed training and model optimization.
Execution: Ability to turn ambiguous business problems into modeling solutions and own iteration from offline experiments to online A/B results.

Technologies

Python
C++
PyTorch
TensorFlow

Benefits

Day one access to medical, dental, and vision insurance
401(k) savings plan with company match
Paid parental leave
Short-term and long-term disability coverage
Life insurance
Wellbeing benefits
10 paid holidays per year
10 paid sick days per year
17 days of Paid Personal Time (prorated upon hire with increasing accruals by tenure)
Additional discretionary bonuses/incentives (eligible may apply)
Restricted stock units (eligible may apply)

About the Team

Live commerce is described as the hardest recommendation problem at TikTok.
The Global E-commerce Recommendation Live Algorithm team owns the end-to-end recommendation stack for TikTok Shop LIVE, from retrieval and pre-ranking to ranking, mixed ranking, and long-term value modeling.
A unified stack serves seven core surfaces, covering public discovery and private follower relationships.
The team is rebuilding the stack around foundation models, including:
- Large recommendation models following scaling laws
- Generative retrieval on multimodal semantic IDs
- Multimodal LLMs connecting content and commerce
- Reward models and RL post-training aligned to long-term value
- AI agents that tune, debug, and evolve the system
What the team ships directly drives GMV, buyer growth, and creator success for TikTok Shop in the US and globally.

Location

Seattle, WA (onsite)

Compensation

USD 207,480 - 368,220 per year
Base salary range in the selected city: $207480 - $368220 annually
Employees may be eligible for additional discretionary bonuses/incentives and restricted stock units
Benefits may vary depending on the nature of employment and the country work location

Job Details

Job code: A254886
For Pay Transparency

Preferred Qualifications

Experience in e-commerce or live-streaming recommendation, ads ranking, or other real-time conversion-driven systems
Hands-on experience with large recommendation models or generative recommendation, including scaling laws, MoE, long-sequence modeling, pre-training, or distillation
Experience in LLM/MLLM post-training (SFT, reward modeling, RLHF, DPO, GRPO, or other RL-based alignment), ideally for recommendation or ranking
Experience building LLM agents (tool use, planning, automated experimentation, or parameter tuning) or applying reinforcement learning to multi-step decision-making like traffic allocation
Experience with cross-domain transfer, LTV/long-term value modeling, uplift modeling and causal inference, or multi-objective optimization
Experience with large-scale training and inference efficiency (mixed precision, model parallelism, GPU serving optimization)
Publications at top venues such as NeurIPS, ICML, ICLR, KDD, SIGIR, WWW, RecSys, ACL, or CVPR, or strong results in major ML competitions

Accommodation

TikTok provides reasonable accommodations in recruitment for candidates with disabilities, pregnancy, sincerely held religious beliefs, or other protected reasons.
Request assistance here: https://tinyurl.com/RA-request

About TikTok

TikTok is described as the leading destination for short-form mobile video.
Global headquarters: Los Angeles and Singapore.
Additional offices: New York City, London, Dublin, Paris, Berlin, Dubai, Jakarta, Seoul, and Tokyo.

Los Angeles County (Unincorporated) Fair Chance

Qualified applicants with arrest or conviction records will be considered in accordance with federal, state, and local laws, including the Los Angeles County Fair Chance Ordinance for Employers and the California Fair Chance Act.
Criminal history may have a direct, adverse, and negative relationship on certain job duties and could result in withdrawal of a conditional offer.
Role may involve interacting and occasionally having unsupervised contact with internal/external clients or colleagues.
Role requires appropriately handling confidential information and access to IT systems.
Requires exercising sound judgment.

View full job page →

1mo ago

Machine Learning Engineer, Human Centered AI - Evaluations & Insights

Apple Seattle, WA

Seattle, WA

1mo ago Closed

$142k - $263k/yr Full time Adversarial Ai Evaluation Ai Evaluation Artificial Intelligence

This position is no longer accepting applications (closed Sep 29, 2026).

Job Description

Join Apple’s Human-Centered AI team to build evaluation and insight systems that help foundation models and generative AI deliver reliable, safe, and human-aligned experiences. This role focuses on bridging human perception with model performance through robust evaluation frameworks, scalable MLOps pipelines, and automated analysis that turns findings into actionable guidance for engineering and training.

Location: Seattle, WA (onsite)

Compensation: USD 142,300 - 263,300 per yearly

What you’ll do

Architect and execute comprehensive evaluation suites for LLMs and multimodal models, surfacing edge cases across multi-step reasoning, factuality, adversarial robustness, safety, and alignment.
Build deterministic, heuristic, and LLM-assisted evaluation frameworks (including LLM-as-a-judge and reward modeling) to quantify human-perceived quality metrics such as helpfulness and hallucination rates.
Convert qualitative failure modes into quantifiable loss patterns, programmatic guardrails, and actionable data-mixture adjustments for model training and inference.
Partner with engineering teams to improve model behavior using evaluation telemetry to guide prompt engineering, Retrieval-Augmented Generation (RAG) strategies, and model fine-tuning.
Apply advanced ML methods (including embedding-based clustering, representation learning, and perturbation analysis) to map error taxonomies and latent failure manifolds in model outputs.
Develop robust MLOps workflows that codify evaluation metrics, automate regression testing across model checkpoints, and integrate human-centric assessments into ML CI/CD pipelines.
Design scalable, distributed inference and processing pipelines (such as Ray and vLLM) to run high-throughput evaluations, automated annotation, and output analysis at scale.
Define quantitative evaluation frameworks that capture nuanced human factors, including trust calibration, conversational state tracking, and interpretability.
Build automated evaluation pipelines using LLMs to assess outputs at scale, optimizing for high correlation with human baseline annotations.
Collaborate across Apple teams, including software engineering, product, research, and responsible AI, to translate product needs into scalable and efficient evaluation infrastructure.