{"id":478325,"url":"https://alion.io/job/kog-labs-research-engineer","title":"Research Engineer - LLM Architecture","company":{"id":185339,"name":"Kog Labs","domain":"kog.ai","url":"https://alion.io/company/kog-labs","size_band":"11-50","is_staffing_agency":false,"employer_type":"direct","is_intermediary":false,"listed_via":null,"ats_vendor":"Ashby","truth_index":null},"role":"AI/ML","role_family":"AI/ML","seniority":null,"employment_type":"full_time","work_mode":"hybrid","remote_scope":null,"remote_scope_basis":null,"remote_working_hours":null,"hiring_geo_confidence":"explicit","locations":["Paris, France"],"countries":["FR"],"hiring_countries":[],"hiring_countries_total":0,"salary":null,"salary_estimate":{"min_usd":56000,"max_usd":121000,"period":"year","method":"role_country_seniority_unknown","sample_n":121},"experience_years_min":null,"visa_sponsorship":false,"relocation_package":false,"has_equity":false,"technologies":[{"name":"AI Agents","optional":false},{"name":"LLM","optional":false},{"name":"Mixture of Experts","optional":false},{"name":"Quantization","optional":false},{"name":"Speculative Decoding","optional":false},{"name":"Transformers","optional":false}],"status":"live","first_seen_at":"2026-06-09T13:44:17Z","employer_posted_date":"2026-06-09","last_verified_at":"2026-10-28T09:45:52Z","board_verified":true,"closed_at":null,"days_open":141,"trust":{"level":"stale","repost_count":0,"flags":["stale"],"days_open":140},"description":"About Kog\nKog builds a co-designed inference stack for real-time AI agents on standard NVIDIA and AMD datacenter GPUs.\nWe develop model architectures and inference systems together. Our work includes Laneformer 2B and Delayed Tensor Parallelism (DTP), a Transformer architecture that overlaps communication with useful computation and weight streaming.\nOur inference stack generates 3,500 tokens/s per request on 8 AMD MI300X GPUs and 2,100 tokens/s per request on 8 NVIDIA H200 GPUs, in FP16 at batch size 1, without quantization or speculative decoding.\nOur next challenge is to accelerate how new AI models are adapted and optimized for different GPU architectures.\nOn the research side, we explore architectural innovations, model morphing, and approaches that make models more efficient at inference time. We also plan to build internal research agents to help generate ideas, run experiments, and evaluate architectural changes faster.\nWe're a team of 10 people, including three PhD holders and one PhD candidate.\nTry our technology at playground.kog.ai and explore our research on the Kog Labs blog.\nWhat you will work on\nYou will research and develop model architectures designed around inference constraints, including:\nNew approaches to attention, routing, residual structures, and parallelism.\n\nExtensions to Laneformer 2B and DTP.\n\nArchitecture choices for large MoE models, including communication and execution at inference time.\n\nModel morphing, modifying the structure of trained models while reusing learned weights.\n\nExperiments that connect architectural hypotheses to measurable changes in model quality and inference performance.\n\nInternal research agents to accelerate architectural exploration and experimentation.\n\nYou will work on open research questions, implement your ideas, evaluate their results, and contribute to publications and systems built at Kog.\nWhat we look for\nWe look for researchers who have personally designed or materially changed LLM mechanisms, not only trained or fine-tuned existing models.\nA research doctoral background is required. A completed PhD is preferred, but the thesis does not need to be in LLMs.\n\nStrong understanding of Transformer architectures and their underlying mechanisms.\n\nDemonstrated ability to formulate hypotheses, implement experiments, and interpret results.\n\nExperience owning meaningful research decisions and driving work independently.\n\nStrong implementation skills and clear technical reasoning.\n\nExperience with model morphing is an advantage, not a requirement. Previous experience building AI agents is not required, but you should be interested in contributing to this direction.\nWe review existing research work during the process, including papers, theses, code, or technical projects.\nWhat we offer\nDirect impact on Kog's model architecture and inference technology.\n\nThe opportunity to take research ideas from hypothesis to implementation and real inference measurements.\n\nAutonomy to explore new architectural directions and publish results.\n\nClose collaboration with engineers working across model architecture, inference, and GPU execution.\n\nLocation: Paris, France. Candidates based elsewhere must be willing to relocate to Paris.","description_format":"text","description_chars":3224,"description_truncated":false,"requirements":{"experience_years_min":null,"management_years_min":null,"team_size_min":null,"manages_managers":false,"education":{"level":"phd","optional":false},"security_clearance":false,"languages":[]},"benefits":[],"hiring_locations":[{"name":"France","iso":"FR","kind":"country"}],"hiring_excludes":[],"relocation_offered":true,"industries":["AI Research Labs","AI Compute & Inference"],"lifecycle":[{"event":"open","at":"2026-09-06T15:33:38Z"}],"visa":[],"liveness":{"score":6,"band":"cold","label":"Long shot","p_open":1,"p_active":0.222,"p_room":0.28,"age_days":140,"expected_fill_days":41,"reasons":["conf:0","win:tail","crowd:"],"computed_at":"2026-10-28T05:59:10Z"},"pay":null,"html_url":"https://alion.io/job/kog-labs-research-engineer","json_url":"https://alion.io/job/kog-labs-research-engineer.json","meta":{"generated_at":"2026-10-28T13:51:05Z","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":18802,"day_limit":null,"remaining_today":null,"minute_limit":300,"resets_at":"2026-10-29T00:00:00Z"}},"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":185339},"rest":"https://alion.io/mcp/rest/get_company?id=185339"},"human":"https://alion.io/catalog?offer=company.slices&for=job%2Fkog-labs-research-engineer"},{"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%2Fkog-labs-research-engineer"},{"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%2Fkog-labs-research-engineer"},{"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/kog-labs-research-engineer\", \"note\": \"what you need it for\"}"},"human":"https://alion.io/catalog?offer=company.verify&for=job%2Fkog-labs-research-engineer"}],"agent":{"mcp":"https://alion.io/mcp/jobs","mcp_open":"https://alion.io/mcp/open","next":[{"tool":"get_job","arguments":{"id":478325}},{"tool":"search_jobs","arguments":{"company":"kog.ai","detail":"compact"}},{"tool":"match_people","arguments":{"job_id":478325},"note":"people open to offers who fit this vacancy"}],"note":"The same data through MCP tools: https://alion.io/mcp/open needs no sign-in (compact searches, postings, people); https://alion.io/mcp with an Alion account has everything. Docs: https://alion.io/developers"}}