Join the team revolutionizing AI computing at Tenstorrent. You'll work on TT-Forge, our MLIR-based compiler that enables developers to run AI on all configurations of Tenstorrent hardware using an open-source, performant, and general-purpose compiler. You will be at the forefront of the AI hardware revolution, building compiler technologies that redefine what’s possible.

This role is hybrid based in Belgrade, Serbia.

We welcome candidates at various experience levels for this role. During the interview process, candidates will be assessed for the appropriate level, and offers will align with that level, which may differ from the one in this posting.

Who You Are:

• A passionate software engineer eager to work on compiler technologies and the challenges of AI hardware, whether from compilers, systems, or broader software backgrounds.
• Fluent in C++ and Python, with experience building complex systems that bridge high-level frameworks to low-level execution.
• Excited by compiler optimization and machine learning, with experience in PyTorch, JAX, TensorFlow, or deep systems programming.
• A collaborative problem-solver who thrives in open-source and enjoys working closely with hardware and software engineers.
  
  
  

What We Need:

• A drive to solve novel challenges in AI compilation, from optimizing computational graphs to creating custom dialects and transformation passes.
• Experience or strong interest in MLIR and how modular compiler frameworks connect AI models to advanced hardware.
• Motivation to build technology that impacts the future of AI, knowing your work will enable the next wave of breakthroughs.
  
  
  

What You Will Learn:

• How to build open-source compiler frameworks supporting diverse AI models and workloads, including training and multi-chip scaling.
• Deep expertise in compiler technologies including custom MLIR dialects (TTIR, TTNN, TTKernel) and transformation passes.
• New methods for human-in-the-loop compiler optimization using TT-Explorer, making advanced tuning tools usable by all developers.
• How compiler technology powers Tenstorrent’s mission to deliver affordable, open-source AI computing in a highly competitive space.

Tenstorrent offers a highly competitive compensation package and benefits, and we are an equal opportunity employer.

Join our Inference Server Technologies team, where we develop software that powers state-of-the-art AI inferencing on Tenstorrent’s cutting-edge hardware. Our team builds the layer that works on top of the Tenstorrent ML libraries - designing APIs, deploying workloads, and benchmarking end-to-end inference speed. You’ll help us shape how developers consume and scale model execution on Tenstorrent’s stack.

This role is hybrid based in Belgrade, Serbia.

We welcome candidates at various experience levels. During the interview process, candidates will be assessed for the appropriate level, and offers will align with that level, which may differ from the one in this posting.

Who You Are

• An engineer who enjoys designing modern APIs and improving how ML models are deployed in production.
• Curious about performance gains through techniques like batching, caching, and model parallelism.
• Passionate about clean software architecture and effective abstraction layers.
• Motivated to deliver backend systems that developers trust and rely on.

What We Need

• Backend engineers who enjoy solving performance bottlenecks and scaling infrastructure.
• Experience with web technologies, protocols, and system design.
• Familiarity with Python, Docker, and Linux-based environments.
• Strong coding practices and a clear ability to break down complex problems into high-quality, maintainable code.

What You Will Learn

• How to optimize end-to-end ML inference on custom silicon.
• Strategies for building scalable, reliable software interfaces for real-world AI applications.
• How to shape the experience developers have when using Tenstorrent’s hardware for AI workloads.

Tenstorrent offers a highly competitive compensation package and benefits, and we are an equal opportunity employer.

Tenstorrent is building next-generation AI hardware and software. This role focuses on the user-mode driver and interface layer that enables communication, resource management, and performance across our devices. You will play a key role in defining how our software stack interacts with both internal and third-party hardware.

This role is hybrid, based out of Belgrade, Serbia.

We welcome candidates at various experience levels for this role. During the interview process, candidates will be assessed for the appropriate level, and offers will align with that level, which may differ from the one in this posting.

Who You Are

• Deep systems engineer with 7+ years experience in user-mode drivers or low-level systems programming.
• Strong proficiency in C and C++ with solid understanding of hardware-software interfaces.
• Familiar with Linux kernel concepts and how kernel and user-space components interact.
• Experience working with external customers or partners to define and support driver or hardware interfaces.
• Exposure to safety-critical systems, automotive platforms, or Functional Safety standards such as ISO 26262 is a strong plus.

What We Need

• Design and evolve the user-mode driver layer that provides high-performance APIs for Tenstorrent devices.
• Define driver interfaces and specifications that enable the broader software stack to run across internal and third-party hardware.
• Collaborate with kernel and hardware teams to ensure efficient cross-layer design and performance.
• Drive performance engineering efforts, including benchmarking, profiling, and optimization of APIs.
• Support external integrations by defining hardware-software boundaries and enabling reliable deployment in customer environments, including safety-critical use cases where applicable.

What You Will Learn

• How user-mode and kernel-level drivers are designed for high-performance AI hardware.
• How hardware-software interfaces are defined to support scalable and portable systems.
• How performance is optimized across memory, compute, and communication layers.
• How third-party hardware integrations are enabled through well-defined driver and interface layers.
• How safety and reliability requirements influence modern driver architecture, including in automotive and mission-critical systems.

Tenstorrent offers a highly competitive compensation package and benefits, and we are an equal opportunity employer.

Experienced Technical Recruiter - Serbia

People are at the center of this mission. Building the right team is existential. As an Experienced Technical Recruiter, you will play a critical role in identifying the engineers who define where AI goes next and who gets to benefit from it. This is a high-visibility role serving as the recruiting lead and sole talent owner for a critical Tenstorrent site.

This is not about filling seats. It is about building the infrastructure of intelligence with intention, clarity, and reach.

Who You Are

• Technical Expert: 7+ years of experience hiring for deeply technical roles across AI, ML, systems, hardware, or infrastructure. You know how to skill up fast on complex new specs.
• Owner & Pilot: You think independently and take deep ownership. You don't wait for direction; you seek signal, act on insight, and aren't afraid to pilot new strategies to find talent.
• Strategist: You understand how to map a region and nurture passive talent. Sourcing is your primary lever, not an afterthought.
• Trusted Advisor: A thoughtful communicator with the presence to influence stakeholders and maintain a high hiring bar while driving decisions.
• Principled: Grounded in the belief that who we hire shapes what we build. You have excellent judgment and can be trusted to negotiate fairly and handle compensation issues with nuance.

What We Need

• Site Ownership: Full-cycle ownership of critical searches across our Belgrade engineering and operations teams.
• Proactive Sourcing: Identifying talent beyond obvious pools and engaging passive candidates with authenticity.
• Influence & Partnership: Close collaboration with hiring managers to define roles, prioritize needs, and hold the line on quality.
• Operational Excellence: High-quality screening and candidate experience from first outreach to final offer.
• Data Integrity: Proactive communication with management and a commitment to data accuracy. "Not great with data" is not an option here; you will maintain the required reporting and use data to drive your engine.
• Independent Execution: The ability to align quickly with our vision and mission, making sound decisions on the fly.

What You Will Learn

• How to hire for one of the most ambitious technical missions in AI today.
• What world-class talent looks like across disciplines and how to recognize it early.
• How hiring choices shape organizational capabilities, long-term leverage, and company trajectory.
• How to balance urgency with care, and growth with focus, in a rapidly evolving space.
• How to build a talent engine that ensures AI is more open, inclusive, and broadly beneficial.

Tenstorrent is building next-generation AI compute. The Kernel Development and Optimization team develops the performance-critical kernels that unlock the full capability of our hardware across ML and HPC workloads.

This role is hybrid based out of Belgrade, Serbia.

We welcome candidates at various experience levels for this role. During the interview process, candidates will be assessed for the appropriate level, and offers will align with that level, which may differ from the one in this posting.

Who You Are

• A strong C++ systems engineer with experience writing performance-critical or low-level software.
• Comfortable reasoning about concurrency, synchronization, latency hiding, and compute versus memory trade-offs.
• Data-driven in your approach, using profiling and benchmarking results to guide optimization decisions.
• Effective at debugging complex runtime or kernel-level issues in large codebases.
• Structured thinker who can break down ambiguous performance problems into measurable experiments.

What We Need

• Engineers who can design, implement, and optimize GPU-style kernels such as matrix multiplication, attention primitives, and data-movement operations.
• Clear ownership of performance, from identifying bottlenecks to delivering measurable throughput improvements.
• Contribution to host-side orchestration code and parallelization strategies.
• Development of micro-benchmarks, regression tests, and tooling to ensure correctness and sustained performance gains.
• Close collaboration with compiler, runtime, ML, and hardware teams to integrate kernels into production systems.

What You Will Learn

• The execution model, memory architecture, and performance characteristics of Tenstorrent AI hardware.
• How to write and optimize accelerator kernels outside traditional CUDA-first ecosystems.
• Practical AI-assisted and agentic workflows for kernel generation, debugging, and optimization.
• How to translate performance intuition into rigorous, reproducible engineering results.
• How low-level kernels, compilers, runtime systems, and hardware co-evolve in modern AI platforms.

Tenstorrent offers a highly competitive compensation package and benefits, and we are an equal opportunity employer.