Staff Applied AI Researcher (Agentic Search)
Tavily Royaume-Uni
- Lieu
- Royaume-Uni
- Clôture
- Non renseigné
- Publiée le
- 5 novembre 2026
- Domaine
- Agentic Search
- Zone
- Hors d'Afrique — parrainage de visa (selon l'annonce)
Comment candidater
Compte gratuit : « Candidater » prépare votre CV et votre lettre pour cette offre ; vous envoyez ensuite vous-même votre candidature. Vous pouvez aussi postuler directement depuis l'annonce d'origine.
Description
We are seeking a Staff Applied AI Researcher to join a fast-growing team building an agent-native search platform - the web access layer for AI systems.
You can think of this as Google search, but for AI agents: a system designed for machines, not humans. We are building agentic search, where AI systems actively plan, retrieve, evaluate, and refine information rather than simply returning results. As AI becomes the primary interface to the web, this layer will replace the role of traditional search engines.
We are designing these systems to search and reason over web data in real time, under strict latency and reliability constraints. This means solving retrieval, ranking and passage extraction under entirely new access patterns and at significant scale, with systems operating over constantly changing, unstructured data and serving tens of thousands of production workloads 24/7.
This role comes with ownership over key parts of our applied AI research direction and system design, with a strong expectation of defining new approaches and shipping measurable impact in production. In the first year, you will own a specific part of the search stack, e.g. ranking or iterative retrieval.
What you'll work on:
-
Designing an agent-native search system that handles everything from low-latency single-hop search to complex deep search tasks requiring iterative planning, querying, refinement and reasoning.
-
Training ranking and retrieval models optimized for agent-driven workflows under real world constraints
-
Bringing the latest advances from the research community to our search stack and ship it to production.
Your responsibilities:
-
Drive applied research and technical direction across retrieval and ranking systems
-
Design and evolve multi-stage retrieval architectures (query understanding, rewriting, retrieval, reranking and iterative retrieval)
-
Lead experimentation on modern retrieval approaches (embedding research, sparse and hybrid retrieval, reranking, context compression) and bring them into production
-
Develop methods for grounding LLMs in real-time web data at scale, using trust, quality and freshness signals to surface reliable sources
-
Partner with the evaluation team to define new metrics for agentic systems, where correctness is not reducible to clicks
-
Analyze trade-offs across relevance, latency, and cost, and work closely with engineering to deploy systems that operate at web scale
-
Mentor researchers and engineers and help raise the technical bar of the team
Must haves:
-
8+ years of experience in applied AI, ML, or software engineering
-
Proven track record of shipping ML or AI systems to production at scale
-
Deep experience with search, retrieval, ranking or building RAG-based assistants
-
Strong understanding of modern deep learning, especially transformers and embeddings
-
Experience with LLM-integrated or knowledge-intensive systems
-
Experience designing evaluation frameworks and metrics for ML systems
-
Strong programming skills in Python
-
Top-tier publications (ICLR, ICML, NeurIPS, AAAI, EMNLP, ACL)
-
Ability to operate in a fast-moving, product-driven environment with high ownership and autonomy
Nice to haves
-
Background in agentic AI systems (agents, tool use, autonomous workflows)
-
Experience with RL post-training of LLMs or VLMs
Find Jobs in United Kingdom on Arbeitnow
Modalités de candidature
Connectez-vous pour ajouter cette offre à vos sélections, puis suivre votre candidature.
Également proposée à
Où se trouve le poste
lieu indiqué par la source : London
Fond de carte © les contributeurs d'OpenStreetMap, chargé depuis leurs serveurs. Coordonnées des villes : GeoNames. Confidentialité
Provenance
vue sur le site d'origine le 7 novembre 2026 à 21:03 (heure de Paris)
Nous vérifions régulièrement que l'offre y est toujours publiée.
Informations non fournies par la source
contrat, durée, mode de travail, salaire, référence