- Company Name
- The Economist
- Job Title
- AI Engineer, AI Lab
- Job Description
-
**Job Title:** AI Engineer, AI Lab
**Role Summary:**
Design, develop, and fine‑tune large language models (LLMs) to match a specific editorial voice and to power multimodal generation (text, voice, and images). Build retrieval‑augmented generation (RAG) pipelines, prototype text‑to‑speech (TTS) systems, and own end‑to‑end quality evaluation processes. Work closely with design, product, infra, journalists, and editors to deliver production‑ready GenAI tools and experiments.
**Expectations:**
- Deliver end‑to‑end GenAI solutions from concept to production.
- Collaborate cross‑functionally, providing technical leadership on LLM strategy.
- Establish and maintain rigorous evaluation and feedback loops, including human‑in‑the‑loop assessments.
- Influence editorial standards and metrics for AI‑generated content.
- Advocate best practices in data curation, model safety, and ethical use of AI.
**Key Responsibilities:**
1. Fine‑tune LLMs (OpenAI, Claude, Cohere, Gemini, Mistral, HuggingFace) for style, tone, and editorial alignment.
2. Curate and version training datasets; manage annotation workflows.
3. Design and evaluate RAG pipelines that retrieve structured content.
4. Prototype and test TTS pipelines for audio‑first products using ElevenLabs, OpenAI TTS, Whisper, etc.
5. Build internal demos and tools with infra, frontend, and design teams.
6. Own evaluation pipelines (BLEU, ROUGE, custom editorial scoring); integrate human feedback.
7. Conduct product experiments (real‑time generation, summarisation, personalization).
8. Partner with journalists and editors to create evaluation metrics reflecting editorial nuances.
**Required Skills:**
- 3+ years experience building LLM/NLP pipelines; hands‑on with OpenAI, Claude, Cohere, Gemini, Mistral, HuggingFace.
- Proficient in supervised fine‑tuning, prompt tuning, instruction tuning, and RLHF/RLAIF principles.
- Strong Python skills; experience with LangChain, HuggingFace Transformers, Pandas, DVC, Weights & Biases.
- Familiarity with eval metrics (BLEU, ROUGE) and building custom evaluators.
- Experience with STT/TTS tools (Whisper, ElevenLabs, Bark).
- Excellent communication and collaboration abilities; ability to translate technical concepts for editorial stakeholders.
- Curious, exploratory mindset; comfortable working in ambiguity and pushing GenAI boundaries.
**Required Education & Certifications:**
- Bachelor’s degree in Computer Science, Data Science, Machine Learning, or equivalent practical experience.
- No specific certifications required, but familiarity with open‑source LLM libraries and ethical AI practices is advantageous.