Mid Full-Stack Engineer (React + Python)

Starkdata · Lisbon

About the Role

We are looking for a versatile Data Scientist to lead our AI initiatives, bridging the gap between classical Machine Learning and cutting-edge Generative AI. You will contribute to the development of StarkVision - our autonomous AI agent platform - and maintain our core predictive models for Churn, Segmentation, Forecasting and others. This is a high-impact role where you will not only build models but also architect complex multi-agent systems that interact directly with our users and databases.

Key Responsibilities

  1. Generative AI & Autonomous Orchestration

Agent Orchestration: Design and optimize multi-agent workflows. You will manage specialized agents (Database Admin, Report Agent, Cloud Specialist) to handle complex user requests.

Text to SQL: Enhance our Text-to-SQL capabilities, ensuring accurate translation of natural language into complex SQL queries using LLMs (GPT-4, Gemini, Grok).

RAG & Memory: Maintain and improve our long-term memory systems using Vector Databases (ChromaDB) and RAG pipelines to provide context-aware responses.

Tool Integration: Develop and maintain MCP (Model Context Protocol) tools that allow agents to interact with external services (OneDrive, SQL Databases, PDF parsers).

  1. Machine Learning & Predictive Modeling

Churn Prediction: Maintain and improve our deep learning Churn Model built with TensorFlow/Keras (LSTMs). You will handle 3D time-series data construction and model optimization.

Customer Segmentation: Refine our clustering pipelines using Scikit-Learn and Faiss (for GPU-accelerated clustering). Implement advanced feature selection techniques and hybrid clustering approaches (K-Means + Hierarchical)

Forecasting: Manage time-series forecasting modules for sales and demand prediction.

  1. Backend & Deployment

Production Engineering: MDeploy models and agents within our Flask application, utilizing Redis and RQ (Redis Queue) for asynchronous background processing.

Data Engineering: Write efficient SQL queries and Python scripts (Pandas, SQLAlchemy) to preprocess large datasets for both ML training and agent consumption.

  1. Other

Company Workflows: Assist in requirements definition, making sure everyone is on the same page.

Requirements

• Bachelor's degree in computer science, Software Engineering, Computer Engineering, or related technical field.

• 3+ years of experience in Data Science or Machine Learning Engineering.

• GenAI Expertise: Proven experience building LLM-based applications. Familiarity with Agentic workflows and RAG architectures is a must.

• Deep Learning: Hands-on experience with Neural Networks, specifically LSTMs/RNNs for time-series data (TensorFlow or PyTorch)

• Strong Math/ Stat Foundation: Deep understanding of clustering algorithms, dimensionality reduction (PCA), and statistical forecasting.

• Coding Skills: Expert-level Python. Comfortable writing production-ready code, unit tests, and working with APIs.

• Database Skills: Strong SQL proficiency. You understand database schemas and can optimize queries.

Nice to Have

• Experience with Kubernetes and Helm Charts.

• Knowledge of Cloud Services.

• Experience with Text-to-SQL fine-tuning or prompt engineering.

• Familiarity with Faiss for large-scale similarity search.

• Knowledge of Model Context Protocol (MCP)

• Knowledge of Agent2Agent Protocols (A2A)

• Ability to handle ambiguity in complex requirements scenarios.

• Experience with on-prem solutions.

What we Offer

• A great environment with real world challenges.

• Opportunity to work on high-tech products integrating Analytics, AI and BI for Banking, Retail and Health.

• A modern technical environment (Modern Python, automated workflows).

• Hybrid work format.

More roles at Starkdata