Lead Test-Time Compute & Search Inference Architect

Opportunity Overview

Location: 100% Remote (United States)
Employment Type: Full-Time Direct-Hire via Accel DNA
Compensation Range: $235,000 – $285,000 Base + Equity Options
Hiring Company: Accel DNA – Enterprise Data, Analytics & AI Talent Solutions

Lead the architectural design of test-time search algorithms, monte carlo tree search (MCTS) decoders, and extended reasoning verification loops for frontier autonomous systems.

Key Responsibilities:

• Architect and benchmark dynamic test-time computation policies allowing reasoning models to scale inference latency for mission-critical queries.
• Integrate tree-of-thought and graph-of-thought exploration algorithms into production agentic execution graphs.
• Collaborate with LLMOps teams to profile KV cache memory footprint and throughput on NVIDIA H100/B200 clusters.
• Implement automated process reward models (PRMs) to evaluate intermediate reasoning steps in complex code generation and mathematical proofs.

What You Bring:

• 6+ years in deep learning, transformer architecture, or algorithmic search systems.
• Demonstrated experience with test-time compute scaling, beam search customization, and reinforcement learning.
• Proficiency in PyTorch, C++/CUDA extensions, vLLM, and Triton Inference Server.
• Master’s or Ph.D. in Computer Science, Artificial Intelligence, or Applied Mathematics.

About Accel DNA & How to Apply:

Accel DNA is North America’s specialized talent partner for Data Science, Generative AI, and Machine Learning. Headquartered at 30 Wall Street, New York, we connect top-tier technical practitioners with industry-leading enterprise teams.

Interested in exploring this opportunity? Submit your profile or portfolio directly to contact@accel-dna.com or apply through https://accel-dna.com/contact-us/.