Remote ML Lead: LLM Dialogue Systems & Team Growth

Social Discovery Group; España; On-site; PHP 7,524,000 - 11,285,000; Full time

Benefits offered by this job

Remote role

Vacation 28 days

Wellness days

Bonuses up to $5000 for referrals

Training stipend 50%

English lesson discount

Health benefits up to $1000

Home office stipend up to $1000 every

Job summary

Social Discovery Group (SDG) seeks a Machine Learning Lead to own the ML strategy for dialogue systems and align bets with business metrics. You will manage a team of 3 ML engineers, drive post-training and model adaptation, and design agent harnesses with tool-using LLM systems.

The role requires hands-on work (10–20% time) on prototypes, debugging traces, and ensuring low latency and cost. Fluent Russian and CET-friendly hours are expected.

Qualifications

• Technical degree and ML engineering background; you have trained and shipped models yourself, not only managed people who do.
• Experience leading a team of 3–5 ML engineers, including hiring and performance decisions.
• Expert-level Python, strong understanding of transformer architecture and modern LLM behavior; hands-on with training, fine-tuning and evaluation.
• Production experience with LLM systems: APIs, streaming, batching, fallbacks, cost/latency trade-offs, observability over traces.
• Ability to read fresh research and translate it into prototypes, evaluations and shipped improvements with measurable impact.
• Fluent Russian, ready to work in CET (±2) hours.

Responsibilities

• Own the ML strategy for dialogue systems: decide what to build and in what order, tied to business metrics.
• Lead a team of ML engineers; set the technical bar, distribute work, hire and grow the team.
• Drive LLM post-training and model adaptation end to end: SFT, LoRA/QLoRA, preference optimization, dataset construction.
• Design agent harnesses and tool-using LLM systems: tool calling, structured outputs, routing, retries, memory, guardrails.
• Build the evaluation layer: offline and pairwise evals, LLM-as-a-judge, regression suites tied to AB outcomes.
• Cut dialogue failure modes and keep inference efficient on latency and cost per message.
• Stay hands-on where it matters (10–20%): prototypes, debugging traces, reviewing team work.

Skills

Python

Transformer models

LLM systems

Team leadership

Hiring decisions

Research-to-prototype

Russian language

Education

Technical degree

Job description

Social Discovery Group (SDG) seeks a Machine Learning Lead to own the ML strategy for dialogue systems and align bets with business metrics. You will manage a team of 3 ML engineers, drive post-training and model adaptation, and design agent harnesses with tool-using LLM systems.

The role requires hands-on work (10–20% time) on prototypes, debugging traces, and ensuring low latency and cost. Fluent Russian and CET-friendly hours are expected.