We are seeking a high-impact motivated Software Architect to join our team and help shape the future of our advanced neural network AI Software Toolchain over AI computing processors.

Responsibilities

Lead Software Architecture specification and supervise design for the most advanced Software Toolchain handling state-of-the-art Neural Processing Unit operation.

Collaborate proactively with Product, Architecture, VLSI and Software teams to promote software leadership over various markets and compute SoCs.

Explore and translate state-of-the-art neural network and AI applications requirements into software architecture flows, encompassing hardware, software, tools and other components.

Evaluate architecture proposals, internal and external IP features and provide influential and inspirational leadership across hardware and software to align all parties to a common vision of architecture & technology development.

Represent with high technical credibility in customer meetings, appropriately incorporating feedback.

Boost velocity of development teams by providing technical guidance and by constantly looking ahead to anticipate and resolve future challenges.

Conduct experiments, invent and drive development of supporting tools such as simulators, models, profilers, and other methods as required.

Engage with engineering leadership and product planning stakeholders to develop technology roadmap.

As a senior compiler engineer, you will own the design and development of open source based toolchain and graph compiler. You will drive technical solutions end-to-end, deliver compiler capabilities across all hardware targets, and help mentor engineers. Our team embraces an AI-agentic development flow, leveraging AI coding agents to accelerate engineering productivity, and we are looking for engineers who champion this approach.

Responsibilities:

Design and develop open source based toolchain targeting custom DSP and NPU architectures.

Drive design and delivery of graph compilation flows: transformations, operator mapping, and code generation.

Champion AI-agentic development practices - integrating AI coding agents into day-to-day engineering workflows.

Own complex components end-to-end, from architecture and design through implementation and delivery.

Provide technical guidance and mentorship to engineers; conduct code reviews.

Collaborate with hardware architecture, algorithms, and software teams globally.

Analyze and optimize toolchain output for performance and correctness across processor families.

In this role, you will be a key contributor to the design and implementation of AI Graph Compiler software stack for Neural Processing Units (NPUs). You will take part in defining software architecture, implementing performance-critical components, and enabling efficient execution of advanced neural networks under tight power, memory, and latency constraints.

You will work closely with hardware and system architects, software and hardware engineers, influencing both software and hardware decisions. You will design and implement major parts of NPU embedded solutions, actively promoting AI capabilities to the customers.

What will you do:

Own and design key components of the AI Graph Compiler software stack for NPU-based systems.

Optimize inference performance (latency, throughput, memory footprint, power) for edge deployments.

Collaborate on HW-SW co-design, influencing NPU architecture.

Support IP evaluations and silicon bring-up, root-cause complex HW/SW issues, and influence development methodologies.

Mentor junior engineers and contribute to technical best practices.

In this role, as Embedded Software Team Leader, you will lead a team owning a key part of NPU AI software stack, a complex, multi-layered system spanning the Host (SDK, runtime, APIs, and UMD/KMD drivers) and the NPU IP (embedded firmware). This stack is at the heart of the AI software offering, with deep impact on multiple hardware and software teams ability to execute state of the art LLM, Generative AI and Neural Networks with extraordinary performance. Reporting to the AI Software SDK Senior Team Leader, you will lead a team of creative and result oriented enthusiasts that invent, design, and implement software that enables smooth connectivity between the NPU and Host CPU.



Responsibilities:

Lead and manage the Embedded Software engineering team, ensuring high performance and productivity.

Own the design, development, and implementation of your component within NPU AI software stack, aligned with the AI Software SDK group's technical strategy, architecture, and roadmap and its automated CI and release process.

Collaborate with cross-functional teams to integrate solutions. Provide technical guidance and mentorship to team members.

Ensure project timelines and deliverables are met.

Conduct performance reviews and support professional development of team members.

Collaborate globally with architects, HW teams, NPU teams, and other developer teams

We are looking for a Senior Backend Engineer specialized in Generative AI to design agent workflows, optimize interactions with models (OpenAI, AWS Bedrock), and ensure the reliability of non-deterministic systems in production.


What You'll Do
Agent Architecture: Design and implement complex agent orchestration logic using LangGraph. You will define state management, conditional routing, and error handling within the agent graph.
Tool Engineering: Build and optimize the tool layer (function calling) that allows LLMs to interact with internal financial APIs and databases accurately.
Performance Optimization:
-Reduce end-to-end latency through asynchronous processing and streaming (SSE).

-Implement semantic caching strategies to minimize API costs and response time.

-Optimize token usage without sacrificing answer quality.

Observability & Evaluation: Implement automated evaluation pipelines using LangSmith. You will be responsible for setting up regression testing for prompts and agents to measure quality (correctness, faithfulness) before deployment.
Advanced RAG: Refine retrieval strategies. Work on hybrid search implementation (Keyword + Vector), re-ranking, and query expansion to feed the most relevant context to the model.