Join us in San Francisco.

What you'll do

• Design and train models that predict brain activation (fMRI BOLD signal) from visual and language inputs
• Extend our RL training pipeline using neural reward signals (building on TRIBE v2)
• Run ablations and evaluations across cortical regions, publishing results publicly
• Collaborate closely with the engineering team to integrate research into production systems
• Help define Amphora's scientific roadmap and methodology

What we're looking for

• PhD or equivalent research experience in computational neuroscience, ML, or a closely related field
• Hands-on experience with fMRI data analysis (GLM, ROI analysis, surface projection)
• Strong PyTorch skills; experience with RL or RLHF a significant plus
• Track record of shipping research — papers, open-source work, or internal results at scale
• Ability to move fast: prototype, evaluate, and iterate in weeks, not months

What you'll do

• Build and maintain training pipelines for fMRI prediction and RL fine-tuning at scale
• Design and operate model serving infrastructure with low-latency inference requirements
• Own our evaluation harness — automated benchmarks, regression tracking, quality gates
• Collaborate with research to ship model improvements to production quickly
• Instrument and monitor model performance in production; own reliability

What we're looking for

• 3+ years of ML engineering experience, with at least one production model shipped end-to-end
• Strong Python and PyTorch; experience with distributed training (DDP, FSDP, or DeepSpeed)
• Familiarity with model serving patterns (ONNX, TensorRT, vLLM, or similar)
• Experience with experiment tracking (W&B, MLflow) and reproducible ML workflows
• Comfort working in a fast-moving, early-stage environment with significant ownership

What you'll do

• Write and distribute content that explains our research to a broad creative and technical audience
• Manage and grow Amphora's presence on Twitter/X, LinkedIn, and relevant research communities
• Help build the waitlist and early user base through targeted outreach and partnership ideas
• Assist in producing launch materials for new experiments and product milestones
• Track and report on growth metrics; identify what's working and double down

What we're looking for

• Strong writer — you can explain complex ideas simply without losing the nuance
• Genuine interest in AI, neuroscience, or the intersection of technology and creativity
• Some experience building or growing an online audience, community, or brand
• Comfortable working in a small team where the scope of your role will expand quickly
• Based in San Francisco (or willing to relocate) for the summer

What you'll do

• Run fMRI data analyses and help extend our cortical ROI evaluation framework
• Build tooling to visualise and interpret brain activation maps across experiments
• Assist in designing and running ablations on the language model training pipeline
• Write up findings clearly — internal reports and contributions to public write-ups
• Explore new research directions and present ideas to the team weekly

What we're looking for

• Currently pursuing a degree in neuroscience, cognitive science, computer science, or a related field
• Familiarity with neuroimaging concepts (fMRI, BOLD signal, ROI analysis) — coursework or research experience
• Python proficiency; experience with NumPy, nibabel, nilearn, or similar neuroimaging libraries a plus
• Intellectual curiosity and ability to read and apply recent ML and neuroscience papers
• Available full-time in San Francisco for the summer (June–August)