- Location
- Hyderabad, India
- Type
- Full-time
- Department
- Engineering
- Seniority
- Lead
- Source
- Workday
Description
Lead Engineer- AI Compilers
What we do Our game-changing AI solutions revolutionize what people and businesses can achieve. Ara inference processors combined with our SDK deliver unrivaled deep learning performance at the edge to accelerate and optimize real-time decision making where every millisecond is critical, and power efficiency is a must.
This is what you are responsible for
Architect and implement high-impact compiler subsystems: IR design, lowering pipelines, operator fusion, scheduling, and target code generation for our NPU.
Lead design and deployment of quantization flows, mixed-precision strategies, and numerical-correctness/bit-exactness verification.
Implement and ship production-grade codegen and runtime integrations that are stable, debuggable and performant.
Drive low-level optimizations and collaborate with kernel teams to close the HW/SW performance gap (including micro-kernel design and latency/throughput tradeoffs).
Define and enforce best practices for compiler CI, performance regression suites, profiling, and release gating.
Troubleshoot and root-cause complex performance or correctness issues on real customer models; produce reproducible fixes.
Mentor senior/junior engineers, do design reviews, and raise the team’s technical bar.
Influence product and silicon roadmap by identifying compiler-driven feature opportunities and tradeoffs.
• Represent the company in technical engagements with customers and partners (deep integrations, POCs, performance characterizations). Necessary Qualifications
Deep expertise in ML compilers and toolchains (MLIR, LLVM, TVM, XLA, Glow, or similar).
Proven experience with graph lowering, IR design, operator fusion, scheduling, and target codegen for accelerators.
Strong systems programming skills — expert C++ and solid Python for tooling and infra.
Demonstrated track record of shipping production compiler components used in deployed products.
Excellent debugging & profiling skills with performance toolchains (perf, custom profilers, trace analysis).
Strong communication skills and experience working cross-functionally with hardware architects, kernel teams, and product.
Passion for mentorship and improving engineering practices across teams.
