Job VC

We are looking for a
Python Engineer
with experience in
Machine Learning and Large Language Models to join a short-term project
focused on building an intelligent data matching solution.
The project involves developing a reusable tool that automatically matches parent and child records between Excel/CSV datasets using a combination of rule-based logic, semantic search, and AI-assisted approaches. The solution should be configurable, scalable, and capable of handling new datasets without changes to the core business logic.
Project details:
Engagement:
Full-time for approximately 2 weeks, followed by several iterations of improvements based on client feedback (with a possibility to continue in the next project phase).
Start date:
ASAP.
Responsibilities:
Develop a reusable Python-based tool for parent-child matching between Excel/CSV datasets.
Design configurable data processing pipelines that can be executed locally on the client's environment.
Implement text, code, numeric value, unit, and attribute normalization.
Build candidate retrieval using exact matching, fuzzy matching, and multilingual semantic search.
Develop confidence scoring and validation logic to avoid incorrect matches based solely on textual similarity.
Generate structured output (CSV/XLSX) including match confidence, status, and explanation.
Participate in several improvement iterations based on results from previously unseen client datasets.
Communicate with the client regarding edge cases and continuously improve the matching quality.

Requirements:
Strong commercial experience with Python.
Experience building Machine Learning or NLP solutions.
Hands-on experience working with LLMs.
Experience with semantic search, embeddings, or retrieval systems.
Familiarity with fuzzy matching techniques and text preprocessing.
Experience working with structured datasets (CSV, Excel, pandas).
Ability to design configurable, reusable software rather than one-off scripts.
Strong analytical and problem-solving skills.
Good English level for communication.
Nice to have:
Experience with multilingual NLP.
Knowledge of vector databases or embedding-based retrieval.
Experience integrating AI services as fallback mechanisms.
Experience designing configurable CLI/data-processing tools.
What we offer:
Short-term, high-impact AI/ML project with the possibility of extension.
Flexible remote collaboration.
Opportunity to work on an interesting real-world data matching problem combining ML, LLMs, and traditional search techniques.
Direct communication with the client and the ability to influence the final solution.
Competitive hourly compensation based on your experience.