Role Overview :

We're hiring a Generative AI Engineer to build, deploy, and optimize multimodal AI services across text, speech, and vision.

You'll work on RAG, synthetic data generation, agent workflows, and integrate STT/TTS/OCR with scalable backend systems.

Key Responsibilities :

- Generative Pipelines: Design applications for RAG, CAG, text classification, summarization, image/video generation, OCR, and synthetic data generation.

- Multimodal Integration: Work with STT, TTS, IVR, OCR, and vision inputs to enable seamless AI interactions.

- AI Agent Workflows: Develop modular, multi-step orchestrations for document, conversational, and data-based user journeys.

- Containerization & Deployment: Collaborate with DevOps to containerize services, manage Kubernetes orchestration, and implement CI/CD for agile delivery.

- Observability: Instrument services using OpenTelemetry, Prometheus, and logging tools to ensure SLO-driven production reliability.

- Collaboration: Work cross-functionally with product, data science, and frontend teams to define APIs (REST/GraphQL) and ensure smooth integration.

- Documentation & Mentorship: Participate in architecture reviews, write clear documentation, and mentor junior engineers and interns.

Requirements.

- Bachelor's/Master's in Computer Science, Data Science, IT, or related field.

- 2-3 years of experience building AI/ML products in Python.

- Must be proficient in AI-first coding tools like Claude Code, Cursor, Roocode, etc.

- Proven experience in deploying GenAI applications and agents in production.

- Strong hands-on with vector search, embedding-based retrieval, STT, TTS, OCR/vision.

- Familiarity with Docker, Kubernetes, frontend development, and CI/CD workflows.

- Strong debugging, performance tuning, and cost-optimization skills.

- Excellent communication, teamwork, and mentoring abilities.

Technical Stack.

- Languages & Tools (mandatory).

- Python (pandas, scikit- learn, PyTorch, Tensorflow, etc.) | Git/GitHub | AWS or GCP.

- Generative AI stack (mandatory).

- LangChain, LlamaIndex, transformers, frontier LLMs (OpenAI, Anthropic, Gemini models) and open models (DeepSeek, Qwen, Llama and Phi models).

- Vector stores: FAISS, Pinecone, Qdrant, Weaviate, etc.

- Keyword Index: Elasticsearch, Apache Solr, Typesense, etc.

- Validation frameworks: Pydantic, Instructor, etc.

- LLM Abstraction libraries: LiteLLM.

- Asynchronous or parallel programming: asyncio, joblib, etc.

- API frameworks: FastAPI, Flask, etc.

- FE prototyping: Streamlit, Gradio, etc.

- Agentic AI Frameworks (mandatory, - 1).

- Google Agents Development Kit | LangGraph | OpenAI Agents SDK | PydanticAI.

- Speech & Vision (nice- to- have).

- OpenAI Realtime Voice API/Whisper; ElevenLabs/Smallest.ai TTS; LlamaParse/JinaAI/Mistral OCR.

- Observability & Monitoring (nice- to- have).

- OpenTelemetry | Prometheus | LangSmith | Pydantic Logfire .

- Cloud & DevOps (nice- to- have).