Experience: 2-3 years | Experience Preferred: Build applications using GenAI/LLM APIs & Solid
Python skills

Responsibilities

- Design and build AI-powered features using existing GenAI models (OpenAI, Anthropic,
Gemini, etc.) via their APIs for text, image, and multimodal tasks.
- Fine-tune or adapt lightweight open-source models (e.g., using LoRA/QLoRA, PEFT) for
domain-specific use cases when off-the-shelf models fall short.
- Build robust prompt engineering pipelines, including prompt design, evaluation, and iteration
for consistent, reliable outputs.
- Implement RAG (Retrieval-Augmented Generation) pipelines using vector databases
(pgvector, Pinecone, Weaviate, etc.) to ground model outputs in proprietary data.
- Design and orchestrate multi-step AI workflows/agents (tool use, function calling, chaining) for
complex tasks.
- Manage the end-to-end lifecycle of AI features: data preparation, model/API selection,
evaluation, integration, and deployment.
- Collaborate with backend/frontend engineers and product managers to integrate AI
capabilities into existing systems and products.
- Monitor and optimize model performance, latency, and cost (token usage, API costs,
inference costs for self-hosted models).
- Stay current with the fast-evolving GenAI landscape (new model releases, APIs,



techniques)
and evaluate their applicability to our product.
- Ensure ethical AI practices — fairness, transparency, data privacy, and responsible handling
of user data sent to third-party model APIs

Requirements

- Bachelor’s or Master’s degree in Computer Science, AI/ML, or a related field (or equivalent
practical experience).
- 2–3 years of hands-on experience building applications using GenAI/LLM APIs (OpenAI,
Anthropic, Google, etc.).
- Strong Python skills, with experience calling and orchestrating AI APIs in production systems.
- Practical experience with prompt engineering, structured outputs, and function/tool calling.
- Experience fine-tuning or adapting pre-trained/open-source models (Hugging Face
ecosystem, LoRA/QLoRA/PEFT) for specific tasks.
- Experience building RAG pipelines — embeddings, vector search, chunking strategies,
retrieval evaluation.
- Familiarity with agent frameworks/orchestration (LangChain, LangGraph, LlamaIndex, or
similar).
- Understanding of transformer architecture fundamentals — enough to reason about model
behaviour, limitations, and fine-tuning trade-offs (deep training-from-scratch expertise not
required).

📌 AI Engineer (GenAI Applications) (India)
🏢 isummation technologies
📍 India