Logistics Jobs in Tel-Aviv, Israel (Now Hiring) — 17 open

Wiliot was founded by the team that invented one of the technologies at the heart of 5G. Their next vision was to develop an IoT sticker, a computing element that can power itself by harvesting radio frequency energy, bringing connectivity and intelligence to everyday products and packaging, things previously disconnect from the IoT. This revolutionary mixture of cloud and semiconductor technology is being used by some of the world’s largest consumer, retail, food and pharmaceutical companies to change the way we make, distribute, sell, use and recycle products.

Our investors include Softbank, Amazon, Alibaba, Verizon, NTT DoCoMo, Qualcomm and PepsiCo. 

We're looking for a senior individual contributor to be the technical anchor of our GenAI team — someone who has repeatedly shipped generative AI products to real external customers in production, not internal demos or proof-of-concepts. You'll set the bar for how we build, evaluate, and operate AI systems at scale: agents, copilots, RAG and knowledge systems, and the SDKs that expose them to developers.

Responsibilities:

• Design, build, and operate generative AI applications — agents, agentic workflows, copilots, chatbots, and RAG/knowledge systems — that run reliably in production at scale.
• Own the full lifecycle: retrieval and context design, prompting, tool/function calling, evaluation, monitoring, failure analysis, and resilience.
• Treat evaluation and observability as first-class engineering — not an afterthought. Define how we measure quality, catch regressions, and debug failures (hallucination, context overflow, broken tool calls, drift).
• Build secure, cost-aware architectures for LLM-driven services, with sensible fallbacks, output validation, and guardrails against misuse (e.g. prompt injection).
• Contribute to SDKs and APIs that let external developers build on our AI capabilities.
• Turn Wiliot's ambient IoT data into practical AI-driven experiences across logistics, sustainability, and inventory.
• Partner closely with software engineers, product, and business stakeholders — and raise the technical bar for everyone around you.

Responsibilities

null

Requirements

• 5+ years of software engineering experience, with a meaningful stretch of it spent building and shipping GenAI products.
• A track record of GenAI systems running in production for external customers — you can point to things real users depend on, and talk concretely about what broke and how you fixed it.
• Deep, hands-on understanding of how modern GenAI systems actually work in production:
• Agents & orchestration — building agentic workflows, tool/function calling, memory and state management, and reasoning about when not to use an agent.
• Retrieval & knowledge systems — chunking strategies, embeddings, vector search, hybrid retrieval, and how to evaluate retrieval quality (not just wire it up).
• Evaluation & monitoring — prompt/version management, grounding and quality evals, LLM-as-judge, real-time observability, and systematic failure analysis.
• Resilience & safety — output validation (schema/type enforcement), fallback logic, and defending against prompt injection and other failure modes.
• Deployment economics — caching, model routing, and cost/latency/quality tradeoffs at scale.
• Strong general engineering fundamentals (Python and modern service/infra tooling — e.g. FastAPI, Docker, Kubernetes — or equivalent in your stack of choice). We care about your judgment and what you've built, not which framework you used.
• Clear communicator who can move fluidly between technical depth and business context, and thrives in a fast-paced, ambiguous environment.

Advantage:

• Experience with real-time, streaming, or IoT/ambient-sensing data.
• Built developer-facing AI SDKs or platforms.
• Background in NLP, multimodal AI, or knowledge graphs.
• Working knowledge of data science fundamentals (you don't need to be a DS, but you should be comfortable around them - it will be a big plus).
• Familiarity with autonomous agent frameworks and agent-evaluation tooling.

#LI-Hybrid