AI Agent Engineer

B.E/B.Tech/M.Tech in Computer Science, AI/ML, Linguistics, or equivalent
Strong Python programming skills and experience with ML frameworks
Ability to work with large-scale conversation data
Excellent analytical and problem-solving skills
Strong verbal and written communication skills and ability to explain technical concepts clearly
Willingness to travel in the US for customer engagements
Proficiency in prompt engineering techniques (few-shot learning, chain-of-thought meta-prompting, etc.)
Knowledge of SQL and data analysis tools (Pandas, NumPy)
Experience with experiment tracking and MLOps tools
Understanding of real-time system constraints and latency optimization

Design, test, and refine prompts for voice AI agents handling complex financial conversations
Develop systematic prompt engineering methodologies and best practices for voice interactions
Ability to meta-prompt and run experiments with different LLMs, bringing in new agent architectures to bring in the lowest latencies possible while maintaining high accuracy
Create prompt templates and frameworks that scale across different use cases and customer segments
Implement A/B testing strategies to measure and improve prompt effectiveness
Analyze conversation transcripts and performance metrics to identify improvement opportunities
Use our Simulation Platform and existing call corpus of our customers to tune and improve AI Agent performance
Develop automated evaluation frameworks for AI agent quality assessment
Create feedback loops (semi or fully automated) between production data and prompt refinements
Optimize prompts for latency, accuracy, and natural conversation flow
Collaborate with ML engineers to improve voice recognition and synthesis quality
Design conversation flows that handle edge cases and error recovery gracefully
Implement context management strategies for multi-turn conversations
Develop domain-specific language models for financial services
Travel to customer sites to learn from human agents and gather requirements
Conduct on-site prompt optimization based on customer-specific needs to deliver rapid iterations of prototype versions