6 months ago
Hyderābād, IndiaMid Level
Responsibilities
- Design, develop, and implement GenAI solutions using large language models and Python.
- Collaborate with stakeholders to identify GenAI opportunities and translate requirements into scalable solutions.
- Preprocess and analyze unstructured text and documents for model training, fine-tuning, and evaluation.
- Apply prompt engineering, fine-tuning, and RAG techniques to optimize LLM outputs.
- Deploy GenAI models and APIs into production while ensuring performance, scalability, and reliability.
- Monitor and maintain deployed solutions and improve them based on feedback and real-world usage.
- Stay current with GenAI, LLM, and orchestration-tool advancements.
- Write maintainable, documented code and contribute to code reviews and team best practices.
Requirements
- At least 2 years of relevant experience as an AI Developer.
- Proficiency in Python.
- Understanding of GenAI models including OpenAI, Llama 2, and Mistral.
- Ability to set up local GPTs using Ollama, LM Studio, or similar tools.
- Experience with LLMs, RAG, and vector databases such as FAISS and Pinecone.
- Experience with multi-agent frameworks for creating workflows.
- Experience with LangChain or similar tools such as LlamaIndex and LangGraph.
- Knowledge of machine learning frameworks, libraries, and tools.
- Experience with at least one cloud platform: AWS, GCP, or Azure.
- Strong problem-solving, communication, teamwork, independent-work, and time-management skills.
Benefits
- Hybrid workplace arrangement with flexible work schedules and opportunities to work from home.
- Paid time off and holidays.
- Opportunities to collaborate with a diverse team and work from global centers.
- Learning and Development programs, communication and stress-management training, professional certifications, and technical and soft-skill training.
- Family medical insurance, personal accident insurance, health-awareness programs, and extended maternity leave.
- Annual performance bonuses and referral bonuses.
