Generative AI, LLMs & RL

Foundation models, fine-tuning pipelines and reinforcement learning systems: the frontier of applied AI research. Part of our Machine Learning & Generative AI discipline, we work with researchers and engineers building generative systems from early-stage labs through to production at scale.
Lightbulb cube
Job Description

Hiring in Generative AI, LLMs & RL isn't like other technical hiring.

Generative AI candidates worth hiring are usually already deep in a role, publishing at the field's leading conferences, or fielding offers before they've updated a CV. Generic outreach doesn't reach them, and generic recruiters don't know how to qualify them.

At Enigma, we track the researchers and engineers behind the models and papers being published, so when a search opens up, we're not starting from a keyword search.
  • NeurIPS
  • ICML
  • ICLR
  • ACL
  • EMNLP
Many of the teams we support here are also hiring across Computer Vision and ML Infrastructure, so we're used to building the fuller picture of a team, not just filling a single seat.

What a Good Partnership Looks Like

We work best as a close, informed extension of your team, not a CV-forwarding service.
one
Define the Brief
Align on what "great" actually looks like, whether that's a first LLM engineer or a reinforcement learning research lead.
two
Map the Market
Understand who's active in generative AI and RL right now, across foundation model labs, applied startups and research groups.
three
Focused Search
Target the right people, not more people, drawing on relationships built at NeurIPS, ICML and ICLR.
four
Guide the Process
Keep momentum, clarity and alignment from first call through to offer stage.
five
Deliver & Refine
Secure the hire, then refine the approach for next time so every search gets sharper.

Trusted across the ecosystem

FAQs

LLM engineering often blends applied ML with software engineering fundamentals like serving infrastructure, evaluation pipelines and fine-tuning workflows. We assess for that specific blend, rather than treating it as just a subset of general ML hiring.

Author order and role notes help, but we go further, asking candidates to walk us through their specific contribution and probing on the parts of the work they didn't lead, so you're not relying on a paper title alone.

Often, yes. RL specialists, particularly those working on agents or post-training, are a narrower and even more competitive pool than generative modelling more broadly. We treat these as related but distinct searches with different sourcing strategies.

We build counteroffer conversations into the process early rather than leaving them as a surprise at offer stage, and we coach both sides on what a competitive but sustainable offer looks like, so a placement doesn't unravel a week after it's made.

A production-focused hire needs to show they can take a model from a notebook to something serving real traffic reliably, a different skill set from research novelty. We screen for it directly rather than assuming a strong publication record covers it.

Not ready to submit a brief?

Let's just talk it through. Tell us what you're building and we'll help you work out what the hire actually looks like.
2026 Enigma. All Rights Reserved.