Key Responsibilities
● Design scalable Python backends (FastAPI/Flask/Django) to power AI-driven
test execution, integrating Vision LLMs for screen understanding and
locator-free interactions.
● Fine-tune and deploy LLMs/VLMs (e.g., for OCR, layout analysis, dynamic UI
handling) using PyTorch/TensorFlow, with prompt engineering for robust
agent behavior.
● Build production pipelines: data preprocessing, inference, feedback loops,
RAG for test knowledge, and API endpoints for APK uploads/execution.
● Productionize models into real-time workflows on AWS/GCP—containerize
with Docker/K8s, optimize for cost/speed, monitor performance.
● Collaborate on agentic features: reliable execution, precision debugging,
cross-device sync—turning raw screens into actionable insights.
Must-Haves
● 4-6 years in AI/ML + backend (Python expert; FastAPI/LangChain a plus).
● Proven LLM fine-tuning (LLMs/VLMs for vision/NLP; OpenAI/Claude toolkits).
● Backend chops: REST APIs, SQL/NoSQL (MongoDB/DynamoDB), scalable
systems.
● Deployed AI to prod (MLOps: Docker, cloud infra, CI/CD).
● Crisp thinker/communicator—quick to iterate on assignments.