Research Engineer

About the Role

Deloitte has a new AI-first effort, backed by $1B in committed investment, building reasoning models and agentic systems to rebuild how the healthcare system decides across payers, providers, and life sciences. This is a ground-up rebuild of the decision-making machinery behind American healthcare, at national scale, with resourced post-training at scale. As a Research Engineer on the post-training team, you will design, train, evaluate, and align models that reason about healthcare, working across the full post-training lifecycle. Healthcare decisioning is a verifiable-reward domain, grounded in real signals such as clinical policy, adjudicated outcomes, and clinical-expert judgment. You will own the post-training stack for clinical reasoning models end-to-end. This is not a prompt-engineering role; the focus is on improving and shaping model behavior through advanced post-training. A healthcare background is not required; engineers are paired with clinical and domain experts. Hiring is based on demonstrated depth and judgment, not years of experience.

Key Responsibilities

• Post-training & Alignment

• Design and execute post-training pipelines: supervised fine-tuning (SFT), preference optimization, and reinforcement learning / alignment workflows.

• Build and optimize training using techniques such as SFT, RLHF, PPO, DPO, GRPO, RLAIF, and Constitutional AI, and understand their effects on reasoning quality, safety, latency, cost, and reliability.

• Train reasoning models for healthcare decisioning using verifiable-reward RL, designing reward signals and verifiers grounded in clinical guidelines, policy, criteria, and adjudicated outcomes.

• Reward Modeling & Data

• Develop reward models and preference datasets to improve reasoning quality, factuality, safety, policy adherence, and task performance.

• Curate, clean, synthesize, and evaluate large-scale instruction, preference, and domain-specific datasets, with rigorous filtering, deduplication, and quality control.

• Build verification and reward pipelines from proprietary clinical, claims, and operational data and from clinical-expert labeling, turning guidelines, policy, and adjudicated outcomes into checkable reward signals at scale.

• Efficient Fine-tuning, Training & Inference Infrastructure

• Implement efficient fine-tuning strategies including LoRA, QLoRA, PEFT, and adapter-based approaches.

• Build scalable distributed training using DeepSpeed, FSDP, Megatron-LM, Ray, or equivalent.

• Optimize inference performance (latency, throughput, quantization, deployment efficiency) for production, including frameworks such as vLLM, TensorRT-LLM, or TGI.

• Small Language Models & Open-Weight Models

• Train and optimize open-weight models such as Llama, Qwen, Mistral, or DeepSeek.

• Build specialized small language models (SLMs) for on-premise and cloud-hybrid deployment with strong performance-per-dollar.

• Evaluation, Safety & Red Teaming

• Design evaluation frameworks covering reasoning, hallucination detection, factuality, instruction following, structured outputs, and domain-specific metrics.

• Build healthcare-grade evaluation, including held-out clinical benchmarks, deployment regression gates, calibration and uncertainty, factuality against ground truth, and bias/fairness evaluation, co-designed with clinical experts.

• Apply PHI/HIPAA-aware data handling and produce model documentation suitable for regulated clinical use.

• Perform red teaming and adversarial testing to identify alignment failures, unsafe behaviors, jailbreak vulnerabilities, and regression risks.

• Collaborate with agentic and application teams to improve tool use, grounding, and long-horizon reasoning.

Requirements & Tech Stack

• Post-training & Alignment: SFT, preference optimization, reinforcement learning, RLHF, PPO, DPO, GRPO, RLAIF, Constitutional AI.
• Models: Small Language Models (SLMs), open-weight models (Llama, Qwen, Mistral, DeepSeek).
• Efficient Fine-tuning: LoRA, QLoRA, PEFT.
• Distributed Training: DeepSpeed, FSDP, Megatron-LM, Ray.
• Inference Optimization: vLLM, TensorRT-LLM, TGI, Ollama.
• Core Technologies: PyTorch, Python.
• Underlying Concepts: Transformer architectures, tokenization, attention mechanisms, decoding strategies, model scaling trade-offs.
• Data & Evaluation: Large-scale and synthetic datasets, filtering, deduplication, quality control, reward modeling, verifiable reward (RLVR), benchmarking.
• Practices: Production-grade software practices, PHI/HIPAA-aware data handling, safety, governance, responsible AI.

Qualifications

• Education: Bachelor's degree in Computer Science, Machine Learning, Artificial Intelligence, Applied Mathematics, Computational Linguistics, or a related field.
• Experience:
• Demonstrated depth training and post-training large transformer-based language models in production or research. Genuine depth including SFT and at least one preference-optimization or RL method, evidenced by shipped models, releases, or research.
• Hands-on experience with reasoning-model training and/or verifiable-reward (RLVR) workflows.
• Strong understanding of modern post-training techniques.
• Experience with open-weight foundation models.
• Strong expertise in PyTorch and modern deep-learning tooling.
• Experience implementing efficient fine-tuning techniques.
• Deep understanding of transformer architectures and related concepts.
• Strong grasp of LLM evaluation methodologies, benchmarking, reward modeling, and alignment trade-offs.
• Strong Python engineering skills and production-grade software practices.
• Ability to work through ambiguous, highly complex technical problems in fast-moving environments.
• Experience building or optimizing reasoning models, agentic models, or tool-using LLM systems.
• Familiarity with inference optimization frameworks.
• Experience with multimodal models, speech models, or domain-specific foundation models is a plus.
• Experience using large-scale GPU clusters and distributed compute is a plus.
• Contributions to open-source AI projects, research publications, benchmark development, or model releases are a plus.
• Familiarity with safety, governance, and responsible-AI practices is a plus.
• Experience in regulated or high-stakes industries such as healthcare, finance, insurance, or public sector is a plus.
• Travel: Ability to travel 0-50%, on average.
• Sponsorship: Limited immigration sponsorship may be available.

Salary Range

$110,700 - $379,200