Senior Software Engineer, AI Gateway
Kong · Hybrid, Milan, IT
77 180 - 95 340 EUR / Рік · Опубліковано 26 жовт. 2026 р.
Подати заявку на цю вакансію з Jobrona
Створіть акаунт і завантажте резюме. Якщо вакансія вам підходить, ми першою підготуємо й надішлемо саме цю заявку та покажемо, що саме надіслано.
Про вакансію
Are you ready to unlock intelligence?
If you don’t think you meet all of the criteria below but are still interested in the job, please apply. Nobody checks every box - we’re looking for candidates that are particularly strong in a few areas, and have some interest and capabilities in others.
THE MISSION
The AI Gateway team is on a mission to build the central nervous system for the agentic era, enabling developers to securely and reliably connect, observe, and manage their AI and ML services at scale. Some of the gateway’s more advanced capabilities, such as intelligent model routing, semantic caching, request classification, and content guardrails, work best when backed by small, purpose-built models. In this role you will build those features end to end: the gateway code that runs them in the request path, and the models behind them.
WHAT YOU’LL DO
- Own the technical direction for model-backed AI Gateway features, from design to production rollout.
- Build and maintain high-performance components for Kong’s AI Gateway in Rust and Lua.
- Train, fine-tune, and evaluate models for specific gateway use cases, such as classifiers for intelligent routing, embeddings for semantic caching, and guardrail models.
- Curate and maintain the datasets used for training and evaluation, and define the metrics and thresholds that decide whether a model is good enough to ship.
- Integrate models into the gateway request path, optimizing inference for latency, throughput, and memory footprint.
- Troubleshoot bugs, tune performance, and do low-level optimizations across both the gateway and the model serving stack.
- Work with product management and other engineering teams to turn model capabilities into reliable, documented product features.
- Mentor other engineers on the team on ML practices and model integration.
WHAT YOU’LL BRING
THE TOOLKIT
- 3+ years of experience in server-side software development: building and debugging applications that are highly scalable, distributed, reliable and resilient.
- 2+ years of experience in machine learning: building, training and evaluating models for production or research.
- Proficiency with tensor-based deep learning frameworks, in Python and beyond.
- Experience with dataset curation and evaluation methodology.
- Experience running models in production outside of Python, optimizing inference for latency and memory (e.g. ONNX Runtime, quantization, batching).
- Working knowledge of embeddings and vector databases.
- Hands-on expertise with Linux systems.
THE KONG DNA
- Ownership: You take full ownership of your work from concept to production, including the documentation: engineers at Kong write the documentation for their products.
- Pragmatism: You can tell when a model is the right tool and when a simpler solution is, and you measure before you decide.
- Collaboration: You communicate clearly in writing and work effectively with a remotely distributed team.
- Leadership: You make and explain technical trade-offs where there is no established playbook.
BONUS POINTS
- Contributions to open-source software projects.
- Proficiency in Rust, Lua or C.
- Familiarity with the Rust ML ecosystem (e.g. Candle, Burn, tokenizers).
- Experience with MLOps tooling and pipelines for training, versioning, and deploying models.
- Experience with experiment tracking tools (e.g. MLflow, Weights & Biases).
- Experience with LLM post-training techniques such as LoRA or other parameter-efficient fine-tuning methods.
- Knowledge of networking fundamentals and protocols on both L4 and L7 (TCP/IP, HTTP/2, gRPC).
- Hands-on experience with NGINX and OpenResty.
#LI-EA1
About Kong:
Kong Inc., the AI Connectivity Company, is building the connectivity layer of AI. Trusted by the Fortune 500® and AI-native startups alike, Kong’s unified API and AI platform enables organizations to secure, manage, accelerate, govern, and monetize the flow of intelligence across APIs and AI traffic — on any model, any cloud. For more information, visit www.konghq.com http://www.konghq.com.