{"id":137439,"url":"https://alion.io/job/harmonic-research-engineer-training-inference","title":"Research Engineer, Training & Inference","company":{"id":48286,"name":"Harmonic","domain":"harmonic.fun","url":"https://alion.io/company/harmonic-fun","size_band":null,"is_staffing_agency":false,"employer_type":"direct","is_intermediary":false,"listed_via":null,"ats_vendor":"Ashby","truth_index":{"grade":"B","score":75,"open_postings":5,"ghost_share":0,"stale_share":1,"repost_share":0,"time_to_fill_p50_days":null,"computed_at":"2026-10-16T05:45:00Z"}},"role":"AI/ML","role_family":"AI/ML","seniority":"senior","employment_type":"full_time","work_mode":"on_site","remote_scope":null,"remote_scope_basis":null,"remote_working_hours":null,"hiring_geo_confidence":"structured","locations":["Palo Alto, United States"],"countries":["US"],"hiring_countries":[],"hiring_countries_total":0,"salary":{"min":200000,"max":450000,"currency":"USD","period":"year","gross":null,"usd_annual":450000},"salary_estimate":null,"experience_years_min":5,"visa_sponsorship":false,"relocation_package":false,"has_equity":false,"technologies":[{"name":"CUDA","optional":false},{"name":"CUDA Toolkit","optional":false},{"name":"JAX","optional":false},{"name":"LLM","optional":false},{"name":"NCCL","optional":false},{"name":"Python","optional":false},{"name":"PyTorch","optional":false},{"name":"PyTorch C++","optional":false},{"name":"Reinforcement Learning","optional":false},{"name":"TensorFlow","optional":false},{"name":"TensorFlow C++","optional":false},{"name":"C++","optional":true},{"name":"CUTLASS","optional":true},{"name":"FSDP","optional":true},{"name":"Triton","optional":true}],"status":"live","first_seen_at":"2026-05-07T16:38:25Z","employer_posted_date":"2026-05-07","last_verified_at":"2026-10-16T19:29:18Z","board_verified":true,"closed_at":null,"days_open":162,"trust":{"level":"stale","repost_count":0,"flags":["stale"],"days_open":161},"description":"About Harmonic\nAt Harmonic, we are building a mathematical reasoning engine that operates with absolute precision. While most AI makes maximum-likelihood guesses, Harmonic's Aristotle uses Lean 4 and reinforcement learning to verify its reasoning and results.\nFollowing our Gold Medal-level performance on the 2025 International Math Olympiad (IMO) and the successful resolution of long-standing open problems, we are proving that AI can master the most rigorous domains of human thought. Backed by some of the world’s most prominent investors, we are intentionally scaling an elite technical team.\nVisit ourcompany blogto learn more about what we are working on!\nAbout the Role\nWe are developing reinforcement learning systems at a scale where standard abstractions frequently fail. Unlike labs that operate primarily through high-level wrappers, we own the entirety of our RL stack. This ownership spans from low-level environment simulators and custom communication primitives to our distributed training loops and inference engines.\nWe are seeking engineers who view existing libraries as a baseline and the hardware speed itself as the true target. You will be responsible for the architecture powering our agents, with a relentless focus on maximizing the throughput of our reinforcement learning and production workflows.\nKey Responsibilities\nTotal Stack Ownership: Maintain and optimize our proprietary RL training and serving infrastructure. You have the authority to refactor any layer-from the Python API down to the CUDA kernels-to achieve peak performance for foundation model workloads.\n\nOptimized Training: maximize the throughput of our reinforcement learning system from data generation to model training with sharded multi-node training and inference algorithms.\n\nHigh-Performance Serving: optimize our inference stack for high-throughput reinforcement learning and low-latency LLM production traffic. Tune the inference engine, router, and scheduler, down to custom kernels if need be.\n\nCompute Optimization: Identify and resolve performance bottlenecks within our distributed clusters, ensuring optimal throughput and memory efficiency for multi-billion parameter models, balancing memory constraints with compute-heavy training cycles.\n\nMinimum Qualifications\nBS in Computer Science or a related technical field, or equivalent industry experience\n\n2+ years of relevant, hands-on industry experience\n\nProficiency in Python\n\nExperience building or maintaining components within ML frameworks (e.g., PyTorch, JAX, or TensorFlow).\n\nProficiency in either:\nUnderstanding of distributed training concepts and collective communication primitives (e.g., NCCL).\nOR\n\nPractical experience deploying and profiling models on GPU-accelerated cloud infrastructure.\n\nPreferred Qualifications\nMS or PhD in Computer Science, Mathematics, or a related field.\n\n5+ years of relevant, hands-on industry experience\n\nProficiency in C++\n\nExperience writing or improving kernels (Triton, CuTeDSL, TileLang, CUDA, CUTLASS, ThunderKittens) to resolve low-level bottlenecks.\n\nProven success deploying performant inference at scale using open-source or custom inference engines, routers, etc.\n\nDirect experience scaling models via FSDP, Tensor Parallelism, or related sharding techniques on multi-node GPU clusters.\n\nExperience designing reinforcement learning systems for high-throughput training and asynchronous data sampling.\n\nWhat We Offer\nUnlimited PTO\n\n401(k) matching\n\n100% employer-paid health, vision, and dental benefits for employees and 50% coverage for dependents. Harmonic offers varied health coverage options to select what is best for you and your family.\n\nHealth Savings Account (HSA) available for qualifying health plans\n\nEqual Opportunity Statement\nHarmonic is committed to diversity and inclusivity in the workplace. We are an equal opportunity employer and do not discriminate on the basis of race, religion, national origin, gender, sexual orientation, age, veteran status, disability or any other legally protected status.","description_format":"text","description_chars":4038,"description_truncated":false,"requirements":{"experience_years_min":5,"management_years_min":null,"team_size_min":null,"manages_managers":false,"education":{"level":"bachelor","optional":false},"security_clearance":false,"languages":[]},"benefits":["Unlimited PTO"],"hiring_locations":[],"hiring_excludes":[],"relocation_offered":false,"industries":["AI for Science"],"lifecycle":[{"event":"open","at":"2026-05-07T16:38:25Z"}],"visa":[],"liveness":{"score":7,"band":"cold","label":"Long shot","p_open":1,"p_active":0.265,"p_room":0.28,"age_days":161,"expected_fill_days":24,"reasons":["conf:3","win:tail","crowd:"],"computed_at":"2026-10-16T05:45:00Z"},"pay":{"stated_usd_annual":450000,"is_top_pay":true},"html_url":"https://alion.io/job/harmonic-research-engineer-training-inference","json_url":"https://alion.io/job/harmonic-research-engineer-training-inference.json","meta":{"generated_at":"2026-10-16T23:11:51Z","cache_seconds":300,"methodology":"https://alion.io/methodology","terms":"https://alion.io/terms","contact":"https://alion.io/contact","api":"https://alion.io/developers","about":"Alion is a live layer of people, companies and AI agents: who they are, whether they are real and active right now, what they do and how to work with them, readable by people and by agents and paid per call.","catalog":"https://alion.io/catalog.json","usage":{"tier":"crawler_verified","counted_by":"address","units_charged":1,"used_today":5086,"day_limit":null,"remaining_today":null,"minute_limit":300,"resets_at":"2026-10-17T00:00:00Z"}},"offers":[{"id":"company.slices","title":"One company in depth, by slice","status":"live","price":{"credits":0.02,"usd":0.002,"plus_per_slice":{"credits":0.05,"usd":0.005}},"unit":"per company, plus each slice with data","note":"the employer in depth","call":{"mcp_tool":"get_company","arguments":{"id":48286},"rest":"https://alion.io/mcp/rest/get_company?id=48286"},"human":"https://alion.io/catalog?offer=company.slices&for=job%2Fharmonic-research-engineer-training-inference"},{"id":"market.stats","title":"A market slice: pay, demand and time to fill","status":"live","price":{"credits":1,"usd":0.1},"unit":"per slice","note":"pay, demand and time to fill for this role and place","call":{"mcp_tool":"market_stats"},"human":"https://alion.io/catalog?offer=market.stats&for=job%2Fharmonic-research-engineer-training-inference"},{"id":"job.search","title":"Open jobs by role, technology, place, pay and visa","status":"live","price":{"credits":0.02,"usd":0.002},"unit":"per posting in a list","note":"similar open postings","call":{"mcp_tool":"search_jobs"},"human":"https://alion.io/catalog?offer=job.search&for=job%2Fharmonic-research-engineer-training-inference"},{"id":"company.verify","title":"Is this company real and active right now","status":"pilot","price":null,"unit":"per company","request":{"url":"https://alion.io/catalog/request","method":"POST","body":"{\"offer\": \"company.verify\", \"for\": \"job/harmonic-research-engineer-training-inference\", \"note\": \"what you need it for\"}"},"human":"https://alion.io/catalog?offer=company.verify&for=job%2Fharmonic-research-engineer-training-inference"}]}