The role
You will be part of the ML team at Mavenoid, shaping the next product features to help people around the world get better support for their hardware devices. The core of your work will be to understand users’ questions and problems to fill the semantic gap.
The incoming data consists mostly of textual conversations, search queries and documents (more than 1M text conversations per month and growing volume on voice). You will help to process this data and assess new LLM and NLP models to build and improve the set of ML features in the products.
Tech stack
We work:* in Python
with NLP/ML libs, including langchain, langfuse, huggingface, pytorch (among others)
major LLM providers (OpenAI, Anthropic, Google, Mistral) and hosted models
deploying with docker on GCP cloud services
We are pragmatic on which tool to use for each approach, as long as it can be properly packaged for production.
Way of working
We are a small team — by design — and share responsibilities. We care about:* shipping to production and see usage
keeping up with the ML developments
balance between speed and codebase quality
You will:* work fully remote and meet IRL few times a year
focus on specific features and own the process from scoping to production delivery
evaluate ideas and propose the right metrics to explore/implement/ship new things
contribute on ML models and features but also service architecture and the platform at scale
Qualifications
You are an ML engineer who cares about product and user outcomes
At least 4 years of industry experience in ML/data-science roles, specifically in NLP/generative and with conversational data
Experience with ML problem-solving, diagnosing errors and hypothetising next steps
Experience with shipping ML services using Docker (build images, manage revisions), GCP services (cloud run, instances, vertex) and CI/CD practices
Experience with real-time LLM services for RAG conversational systems in production
Experience with voice or agentic system is a plus
Experience with working in a compact ML team with shared responsibilities \& ownership
Responsibilities
scope, build, and deliver ML features to production
thinking ahead for long-term ML development in the product
following software and ML engineering best practices to keep things humming
Day-to-day at the individual level:* 40% exploring/developing ML/NLP problem
10% making sure the ML features are solving the right problem with the right assumptions with product team
30% shipping for production and keeping live features
20% free exploration/investigation for long term
In your first month, you will* Complete Mavenoid’s remote onboarding program
Meet with the ML/Product/CS teams to understand what is being worked on
Familiarize yourself with our platform and product, and processes
Ramp up the codebase with co-working sessions and/or time on your side
Focus on one feature to understand the evaluation metrics and propose a step ahead in accuracy/efficiency/performance
In your 3 months, you will* Work on one first feature improvement to go over explore/implement/evaluate/ship loop
Collaborate with the rest of the team to bring your input on the system architecture and the product
Take over one service and push the envelope
Tackle one or more new feature, from data exploration to feasibility and concept assessment, in collaboration with the product lead.
In your 6 months, you will* Propose, discuss, coordinate and implement your first large platform or architecture change
Be familiar with a large portion of the platform, including the details of our CI/CD/evaluation pipeline, machine learning services, and integrations to external systems
Own a part of the platform and be able to identify areas of improvement.