Synechron - Prompt Engineer

Synechron is looking for an experienced Prompt Engineer to design, develop, evaluate, and optimize prompts and LLM-powered solutions. The ideal candidate should have strong hands-on experience working with Large Language Models (LLMs), Prompt Engineering, NLP/ML applications, Python, and model evaluation frameworks.

You will work closely with data scientists, ML engineers, software engineers, and business stakeholders to improve the quality, reliability, relevance, and performance of AI-driven applications.

Key Responsibilities :

- Design, develop, test, and optimize prompts for Large Language Models (LLMs) to achieve consistent and high-quality outputs.

- Work with LLMs such as OpenAI, Anthropic, Llama-family models, and other generative AI models.

- Develop prompt strategies including zero-shot, few-shot, chain-of-thought, role-based, structured-output, and instruction-based prompting where appropriate.

- Analyze LLM responses and continuously refine prompts based on accuracy, relevance, consistency, and business requirements.

- Build and maintain reusable prompt templates, prompt libraries, and evaluation datasets.

- Collaborate with ML engineers and developers to integrate LLM capabilities into production applications.

- Develop Python-based workflows for interacting with, testing, and evaluating LLMs.

- Work with model inference and orchestration frameworks such as Hugging Face Transformers, LangChain, or similar frameworks.

- Develop evaluation methodologies and benchmarks for assessing LLM performance.

- Define and monitor metrics for accuracy, relevance, factuality, consistency, latency, and other model-performance parameters.

- Conduct A/B testing and comparative evaluations of prompts, models, and configurations.

- Work with data teams on data labeling, annotation, dataset preparation, and prompt-based data augmentation.

- Identify failure patterns, hallucinations, inconsistencies, and other undesirable model behaviors.

- Contribute to improving LLM reliability, robustness, and overall user experience.

- Apply appropriate ML safety, responsible AI, bias mitigation, privacy, and content-policy considerations while designing AI solutions.

- Document prompt strategies, experiments, evaluation results, and best practices.

- Stay updated with developments in Generative AI, LLMs, prompt engineering, model evaluation, and AI agent technologies.

Required Skills & Experience :

- 5+ years of overall experience in Prompt Engineering, NLP, Machine Learning, AI/ML Product Development, or a closely related field.

- Strong hands-on experience with Prompt Engineering and Generative AI/LLM applications.

- Practical experience working with one or more LLM platforms/models such as :

i. OpenAI

ii. Anthropic

iii. Llama-family models

iv. Hugging Face models

v. Other commercial or open-source LLMs

- Strong programming skills in Python.

- Experience with LLM/model inference frameworks such as :

i. Hugging Face Transformers

ii. LangChain

iii. LlamaIndex

iv. or similar frameworks

- Strong understanding of prompt design, prompt optimization, and LLM behavior.

- Experience developing and executing LLM evaluation strategies and benchmarks.

- Hands-on experience with A/B testing, experimentation, and model-performance metrics.

- Understanding of NLP concepts and machine-learning fundamentals.

- Experience with data annotation, labeling workflows, dataset creation, and prompt-based data augmentation.

- Understanding of LLM hallucinations, bias, safety, responsible AI, and content-policy considerations.

- Strong analytical and problem-solving abilities.

- Excellent written and verbal communication skills.

- Experience working with RAG (Retrieval-Augmented Generation) applications.

- Knowledge of vector databases and embedding models.

- Experience with LLM agents and tool/function calling.

- Familiarity with structured outputs, JSON-based responses, and schema-constrained generation.

- Experience with LLM observability and evaluation platforms.

- Knowledge of model fine-tuning, LoRA/PEFT, or instruction tuning.

- Exposure to cloud-based AI/ML platforms such as Azure, AWS, or GCP.

- Experience working in enterprise AI/Generative AI environments.

- Understanding of AI governance, responsible AI, and model-risk management.