Contact Us

Responsibilities:

1. Develop, optimize, and operate large-scale distributed LLM training systems to power AI-driven scientific discovery.
2. Work closely with researchers to build, debug, and maintain post-training and reinforcement learning workflows.
3. Build tools and directly support cutting-edge model training. Replace manual processes through continuous automation, improve internal infrastructure, and enhance R&D efficiency and experience—empowering the company’s scientists, AI researchers, and engineers in their daily work.
4. Partner with product and other engineering teams to bridge the gap between science-capable LLMs and the development of “AI scientists.”

Requirements:

1. Master’s degree or above, with strong engineering skills. Experience creating or maintaining influential open-source projects.
2. Work experience closely related to LLM post-training and RL, with the ability to build and scale the RL infrastructure required for LLMs.
3. Experience training models at the scale of hundreds of GPUs, with hands-on experience in 5D parallelism: data parallelism (DP), tensor parallelism (TP), pipeline parallelism (PP), sequence parallelism (SP), and expert parallelism (EP) for MoE models.
4. Extensive hands-on experience with distributed training frameworks such as Megatron-LM, DeepSpeed, and TorchTitan, with proven experience optimizing training throughput for large-scale mixture-of-experts (MoE) models.
5. Open-minded, straightforward, and collaborative. Passionate about advanced technologies.

Bonus Points For:

1. Work experience at major tech companies’ foundation model teams, leading LLM startups, or cutting-edge research institutes.
2. Unique insights into post-training and RL algorithm frameworks.

Responsibilities:

1. Conduct research on cutting-edge AI algorithms, model architecture design for chemistry and materials science.
2. Collaborate with teams in computational chemistry, data engineering, and software engineering to build data infrastructure and automated modeling workflows centered on chemistry and materials.
3. Partner with internal R&D and business teams to identify and drive the adoption of AI4Science solutions in real-world applications.

Requirements:

1. Hold a Ph.D. in artificial intelligence, computer science, computational chemistry, materials informatics, or a related interdisciplinary field.
2. Deep research or hands-on experience in at least one of the following areas:
   
   • Computer-aided synthesis design and retrosynthesis planning
   • Building and fine-tuning large language models (LLMs) for chemistry
   • Foundation models for predicting molecular or material properties
   • Generative models for molecules and chemical reactions
   • Machine-learned interatomic potentials
   • Benchmarking infrastructure for foundation models
3. Proficient in Python, with strong coding and algorithm implementation skills; familiar with PyTorch.
4. Familiar with common machine learning models and methods (e.g., GNN, Transformer, generative models) and able to build models tailored to specific scientific problems.
5. Strong problem-solving skills, with the ability to translate scientific questions into actionable algorithm and model solutions.
6. Strong interdisciplinary communication skills, with the ability to collaborate effectively across computational, engineering, and experimental teams.

Bonus Points For:

1. Background in computational chemistry or molecular simulation, with familiarity with molecular, reaction, or material data.
2. A track record of publications in algorithms or models, open-source projects, or long-maintained code repositories.
3. Familiarity with models and tools in computer-aided synthesis design (CASP) or retrosynthesis (e.g., ASKCOS, AiZynthFinder, RetroXpert, NeuralSym, GraphRetro).
4. Experience developing large-scale software projects, with familiarity with version control systems (e.g., Git) and code testing practices.

Responsibilities:

1. Participate in the development and optimization of high-performance computational workflows for cutting-edge chemistry and materials research.
2. Collaborate with cloud architects and software engineers to build data infrastructure and automated workflows centered on chemistry and materials.
3. Interface with internal business and technical teams to identify and advance future computational chemistry–driven solutions.
4. Support collaborative projects with partner companies and clients, driving the industrial deployment and advancement of technologies.

Requirements:

1. PhD in computational chemistry or a related field, including but not limited to materials science, chemistry, and chemical engineering.
2. Research experience in at least one of the following areas: heterogeneous catalysis simulations, molecular dynamics simulations, battery electrolyte modeling, organometallic catalysis, reaction thermodynamics/kinetics, or crystal structure prediction and analysis.
3. Proficiency in at least two mainstream simulation packages, such as LAMMPS, GROMACS, PySCF, Gaussian, ORCA, VASP, Quantum ESPRESSO, or CP2K.
4. Strong problem-solving and analytical skills, with the ability to independently conduct scientific research or industrial application studies.
5. Proficient in programming languages such as Python and C++, with solid capabilities in algorithm development and engineering implementation.
6. Strong cross-disciplinary collaboration skills, with the ability to communicate effectively across teams with diverse technical backgrounds, and a commitment to solving key scientific problems through computational approaches.

Bonus Points For:

1. Experienced in the development and deployment of high-throughput computational workflows.
2. Experienced in contributing to the development of computational software, workflows, or platforms within a team.
3. Experienced in AI-driven coding or model development.
4. Experienced in collaborating with experimental teams.

Responsibilities:

1. Lead the team to complete the synthesis, purification, and structural characterization of organic/inorganic small molecules; ensure experimental quality and pipeline progress.
2. Drive reaction optimization and method development: condition screening, route iteration, process feasibility assessment; use data to guide project decisions.
3. Utilize and integrate analytical tools such as UPLC/HPLC, LCMS, NMR, IR, UV-vis, DLS, etc.
4. Collaborate closely with automation engineers and ML scientists—validate model-recommended conditions, analyze trends, diagnose issues, and extract actionable insights.
5. Maintain standardized record-keeping and reporting: manage digital experimental records (ELN/LIMS/database), generate structured data reports, conduct reviews, and deliver periodic summaries.
6. Strictly adhere to EHS and laboratory best practices; contribute to the continuous improvement of laboratory processes and assets.

Requirements:

1. Doctoral degree in chemistry or a related field (or master’s degree with 5+ years of relevant experience).
2. Demonstrated high-quality wet-lab experience with a complete closed loop of synthesis, purification, and analytical characterization.
3. Strong understanding of reaction mechanisms, retrosynthetic logic, and systematic optimization.
4. Proficiency in common analytical techniques: TLC, HPLC/UPLC, LCMS, NMR, IR; ability to independently interpret spectra and troubleshoot issues.
5. Ability to execute reliably within SOP frameworks and on automated platforms, ensuring data consistency and traceability.
6. Proficiency with SciFinder, Reaxys, and ChemDraw for literature search and route/condition scouting.
7. Strong data literacy—with close attention to data quality, batch-to-batch variability, root-cause analysis of anomalies, and structured record-keeping.

Responsibilities:

1. Design and implement scalable LLM agent architectures, covering core capabilities such as task planning, skill invocation, memory, and multi-agent collaboration.
2. Utilize a variety of middleware, databases, OLAP systems, and cloud-native components to ensure high throughput and high availability.
3. Build system-level evaluation benchmarks, and continuously monitor and optimize agent performance across dimensions including accuracy, stability, latency, and cost.
4. Collaborate closely with algorithm, product, data, and business teams to drive the deployment of agents in real-world scenarios.

Requirements:

1. Bachelor’s degree or above in computer science or a related field, with a solid foundation in data structures and algorithms.
2. Strong system design and troubleshooting skills, excellent communication and collaboration abilities, and a self-motivated attitude—committed to continuous improvement and delivering high-quality code in a fast-paced environment.
3. Proficiency in one or more of the following: Python, Go, Java, C++. At least one year of solid backend development experience; experience with large-scale projects is a plus.
4. Familiarity with microservices architecture, the core principles and usage of various databases, and common distributed system middleware.

Bonus Points For:

1. Familiarity with AI or large language models, or a strong curiosity and passion for learning, along with the ability to quickly master new technologies and apply them to real-world business scenarios.
2. Experience with at least one mainstream agent framework (e.g., LangChain, LangGraph, Agnes), including hands-on experience in end-to-end pipelines, evaluation, and tuning.

Responsibilities:

1. Full-Cycle Client Management: Independently acquire and develop top-tier industrial clients—driving the full sales cycle from lead generation and needs discovery to contract signing and payment collection, with a focus on large-scale projects (starting at the million RMB level).
2. Solution Co-Creation and Technical Alignment: Partner with solution architects to analyze client R&D or production pain points, define solution scope, lead technical value validation, and support client-side review processes.
3. Commercial Process Ownership: Lead commercial negotiations, manage procurement workflows, and define contract terms (milestones, acceptance criteria, payment schedules). Take direct ownership of win rates and collection performance.
4. Industry Best Practices: Synthesize reusable sales strategies, client messaging frameworks, and reference cases into actionable methodologies that enable the broader team to replicate success.

Requirements:

1. 5+ Years of B2B Technical Sales Experience: Proven track record of closing project-based deals (specific case examples required). Familiar with the sales motion of complex, solution-oriented offerings.
2. Technical Client Communication: Ability to engage effectively with R&D, process, and technical leaders—understanding technical boundaries and translating product capabilities into clear business value (coding not required).
3. Domain Expertise in at Least One Area: Deep understanding of client profiles, procurement processes, and project approval logic in one or more of the following:
   
   • Semiconductor / packaging materials (photoresists, CMP, encapsulation resins, etc.)
   • New energy materials (electrolytes, cathode/anode materials, etc.)
   • Industrial software / simulation / HPC / lab automation equipment
   • Enterprise software on-premise deployment / industry-specific solutions
4. Structured Sales Capability: Demonstrated ability to assess opportunities, manage sales cadence, and anticipate risks.

Responsibilities:

1. Agent Capability Evaluation & Iterative Strategy
   
   1. Build and continuously refine Agent evaluation test suites — encompassing offline benchmarks, high-value scenario test cases, and real-world feedback examples. Establish a complete pipeline for annotation and test case data accumulation.
   2. Design practical evaluation methodologies — develop a multi-dimensional, scenario‑based, actionable evaluation framework for Agent capabilities that aligns with market/user feedback and product value propositions.
   3. Establish an “evaluation‑driven iteration” strategy — based on test suites and evaluation methods, define layered optimization strategies, translate evaluation results into prioritized action items, and drive the continuous evolution of Agent capabilities.
2. Agent Product Architecture Design & Evaluation Platform Development
   
   1. Participate in product architecture design — be deeply involved in overall product planning and architecture, take ownership of core module feature design, and define and complete evaluation cases and methods.
   2. Build evaluation platform capabilities — design and implement a comprehensive capability evaluation platform. This should cover management of evaluation cases, prompts, test task execution, and results analysis across Agents, workflows, and tools.
   3. Drive key product capability expansion — leverage evaluation results to identify capability gaps and high-value scenarios. Plan expansion roadmaps using a unified capability layering framework, and collaborate with product and R&D teams to develop, evaluate, and launch new capabilities.

Requirements:

1. Bachelor’s degree or above, with 3+ years of product design and management experience. Exceptional candidates with less experience may be considered.
2. Hands-on experience in AI Agent product design and evaluation projects, with end‑to‑end ownership of at least one of the following mainstream AI products: intelligent agent systems / automated workflow orchestration platforms / RAG and knowledge base systems / Agentic model products / Agent evaluation and tuning platforms.
3. Strong product architecture skills — ability to layer and orchestrate tools, models, and data of varying complexity around R&D processes and business scenarios, leveraging compute and model infrastructure. This should be abstracted into stable, batch‑schedulable, reusable, and evolvable capability modules.
4. Solid data feedback and evaluation mindset — capable of designing quantifiable metrics and evaluation methods based on market and user feedback, continuously improving test suites, driving automated regression and online monitoring, and using quantitative metrics to guide stable product iteration.
5. Excellent cross-functional collaboration skills — able to align goals, break down plans, and manage risks and timelines across pre‑sales, marketing, R&D, and delivery teams.

Bonus Points For:

1. Background in chemistry, materials science, drug design, or experience serving R&D scenarios in research or enterprise settings (e.g., ELN, LIMS, chemical database products).
2. Enterprise delivery and cost awareness — experience with compute cost estimation, SLAs, permissions and auditing, and on‑premise/hybrid cloud deployment.
3. Experience designing and commercializing AI products for professional users or vertical domains.