An agent that writes the tests
A LangGraph agent that generates tests automatically (pytest, httpx) and is integrated into the team's pre-deployment testing framework.
- LangGraph
- pytest
- httpx
- Temporal WorkflowEnvironment
- testcontainers
cv_sveva_pepe.txt
Sveva Pepe is an AI/ML Engineer at Almawave Labs, Rome. Since 2025 she designs the AI architecture of an enterprise product built on LLM agents. Before that, from 2022, she extracted entities and relations from judicial documents for the public sector.
relations
Annotated by hand, in the format of the entity- and relation-extraction systems I build.
§1
I'm responsible for the AI architecture: how LLM agents, workflows and services work together to process documents at scale. Follow a document through it:
Large-scale document-processing pipelines turn incoming documents into structured content the rest of the module can work on.
The AI components run as microservices that expose APIs, containerized and deployed on Kubernetes.
Multi-step AI processing is orchestrated as workflows, so every step is tracked, retried and observable.
LLM agents carry out the document analysis, orchestrated with LangChain and LangGraph.
My focus: how the agents are designed and orchestrated.
Results live in a document database, and a cost-tracking system attributes LLM spend to each stage of the pipeline.
I built the stage-level cost tracking, so AI costs stay measurable.
A LangGraph agent that generates tests automatically (pytest, httpx) and is integrated into the team's pre-deployment testing framework.
I coordinate the technical work on the AI side and the key decisions on the AI architecture, and shape the product roadmap together with product management.
§2
Client delivery for public-administration clients: custom solutions on legal text, where language is long, formal and precise.
LLM-based solutions for named entity recognition and relation extraction on judicial documents, including fine-tuning transformer models.
Prompt-engineering strategies designed for legal terminology and complex sentence structures.
Custom evaluation metrics and standardized preprocessing workflows, reused to improve efficiency across multiple government projects.
§3
Product development (2025 – present) · Client delivery (2022 – 2025). Almaviva Group; the company was renamed from Almawave to Almawave Labs in August 2026.
Optimized performance-testing frameworks with LoadRunner, AppDynamics and Instana.
L3DAS project: built a 3D audio dataset and challenge for speech enhancement and sound event localization.
110/110 cum laude, focus on machine learning and NLP. Certificate of Excellent Student, Sapienza, 2023.
110/110 cum laude.
§4
§5
Earlier university projects (NER, semantic role labeling, quaternion networks for 3D audio, reinforcement learning, robotics) are on GitHub.
§6
Based in Rome, open to relocation for on-site roles in Europe.