AI Agent Engineer

Company: Trida Labs
Location: Delhi , India
Work Model: Remote
Employment Type: Full-time
Experience: 3–5 Years

About Trida Labs

Trida Labs builds AI-native products that help organizations work with enterprise data, analytics, and intelligent applications.

Our products include:

TridaPad — An AI-native data collaboration platform for querying, analyzing, visualizing, and managing data across SQL, NoSQL, Big Data, and cloud data sources.
AgentSQL — An agentic AI analytics solution that enables users to interact with enterprise data using natural language, generate SQL, analyze data, create visualizations and dashboards, and derive insights.
TridaPad MCP Server — A Model Context Protocol server that enables AI assistants and MCP-compatible clients to interact with TridaPad and its data capabilities.

About the Role

We are looking for an AI Agent Engineer to design, develop, evaluate, and improve production-grade AI agents for enterprise data and analytics applications.

The role is primarily focused on AgentSQL and agentic analytics, including natural-language-to-SQL, schema understanding, tool orchestration, MCP integrations, context management, agent evaluation, and production reliability.

You will build agents that understand user intent, identify the relevant enterprise data, select appropriate tools, generate and execute analytical workflows, validate results, and communicate insights reliably.

This role sits at the intersection of Generative AI, agent engineering, enterprise data, and software engineering.

What You'll DoAI Agent Development

Design and develop production-ready AI agents for enterprise data and analytics use cases.
Build agents that understand user intent and execute multi-step analytical tasks.
Define agent capabilities, responsibilities, tools, and execution workflows.
Improve agent accuracy, reliability, consistency, and user experience.
Use deterministic tools, APIs, or conventional software approaches when an LLM or agent does not provide meaningful value.

Agent Workflow & Orchestration

Design and implement multi-step agentic workflows for complex analytical tasks.
Develop workflows for planning, task decomposition, execution, validation, and response generation.
Implement state management, conditional flows, retries, fallbacks, and recovery mechanisms.
Build and maintain agent workflows using LangGraph, LangChain, or similar frameworks.
Design reusable workflow patterns that can support different enterprise data use cases.

AgentSQL & Natural-Language-to-SQL

Develop and improve agentic capabilities for AgentSQL and AI-powered analytics.
Build workflows for schema discovery, table and column selection, SQL generation, query execution, data analysis, visualization, and insight generation.
Enable agents to understand relationships between tables, fields, metrics, filters, aggregations, and business terminology.
Evaluate generated SQL for both syntactic correctness and correctness of the resulting data.
Build evaluation datasets containing representative natural-language questions, expected queries, and expected analytical results.
Investigate failures caused by ambiguous requests, incomplete schema context, incorrect joins, filtering errors, or incorrect business-metric interpretation.
Improve SQL generation and analytical accuracy based on evaluation results and real-world usage.

Enterprise Data Access & Safety

Ensure agents operate only on data and tools available to the requesting user.
Incorporate permission and access-control information into agent and query workflows.
Implement validation and safeguards to prevent unsafe or unintended database operations.
Improve handling of ambiguous, incomplete, or unsupported analytical requests.
Consider query complexity, execution time, and resource usage when generating and executing queries.
Investigate cases where technically valid SQL produces incorrect, misleading, or incomplete analytical results.

Tools & Integrations

Design and develop tools that enable agents to interact with APIs, databases, search systems, and application services.
Implement reliable function-calling and tool-execution workflows.
Define tool interfaces, input validation, output handling, and error management.
Improve tool selection based on user intent, available context, and task requirements.
Build reusable tools and integrations across agent workflows.

MCP & AI Integrations

Build AI agent integrations using the Model Context Protocol.
Develop and maintain MCP tools that extend agent capabilities.
Enable AI assistants and MCP-compatible clients to interact with TridaPad and enterprise data capabilities.
Design reliable agent-to-tool interactions with appropriate validation and error handling.
Test, troubleshoot, and improve MCP-based workflows and integrations.

Context & Memory

Design context strategies that provide agents with relevant information during task execution.
Manage conversation state, workflow state, and intermediate results.
Implement memory capabilities where appropriate for multi-step and long-running tasks.
Integrate retrieval systems, embeddings, and vector databases when required.
Improve schema, metadata, and business-context selection to support accurate analytical responses.
Optimize context usage to improve response quality while reducing unnecessary model usage.

Agent Evaluation & Quality

Develop evaluation datasets and test cases for AI agents and NL-to-SQL workflows.
Evaluate agent planning, tool selection, SQL generation, execution, data accuracy, and final responses.
Measure both query correctness and the correctness of analytical results.
Identify recurring failure patterns and investigate their root causes.
Build repeatable evaluation processes for agent quality and reliability.
Improve prompts, tools, context, workflows, and model selection based on evaluation results.

Performance & Optimization

Optimize agent workflows for latency, reliability, and cost.
Reduce unnecessary model calls, tool calls, token usage, and repeated processing.
Select appropriate models based on task requirements and performance considerations.
Optimize SQL generation and execution for complex analytical workloads.
Consider query execution time and resource consumption when working with large enterprise datasets.
Improve agent efficiency while maintaining output quality.

Production & Reliability

Deploy and support AI agents across development, staging, and production environments.
Implement logging, monitoring, tracing, and observability for agent workflows.
Troubleshoot issues across agents, models, tools, APIs, databases, and application services.
Implement appropriate validation, guardrails, retries, and failure-handling mechanisms.
Improve the reliability and maintainability of production agent systems.

Engineering Collaboration

Work closely with a small, cross-functional engineering team across AI, backend, data, platform, and product development.
Translate product requirements into practical agent-based solutions.
Contribute to agent architecture and technical design decisions.
Develop reusable agent components, tools, and workflow patterns.
Document agent architectures, workflows, integrations, evaluation approaches, and known limitations.

What We're Looking For

3–5 years of experience in AI engineering, software engineering, Generative AI, or a related field.
Strong proficiency in Python.
Hands-on experience building LLM-powered applications, AI agents, or agentic workflows.
Strong understanding of LLMs, prompt engineering, and context engineering.
Experience with tool calling, function calling, or workflow orchestration.
Experience working with APIs, databases, and backend services.
Understanding of agent state, workflow execution, and task orchestration.
Experience with natural-language-to-SQL, RAG, embeddings, vector databases, or retrieval systems.
Strong understanding of SQL and relational data concepts.
Strong debugging, analytical, and problem-solving skills.
Strong communication and collaboration skills.

Preferred Qualifications

Experience with LangGraph, LangChain, LlamaIndex, or similar frameworks.
Experience building multi-agent or complex multi-step workflows.
Experience with MCP and AI assistant integrations.
Experience with AI-powered analytics or natural-language-to-SQL systems.
Experience developing evaluation datasets and testing LLM or agent systems.
Experience with AI observability and production monitoring.
Experience optimizing model calls, token usage, latency, and inference costs.
Experience with PostgreSQL, MySQL, MongoDB, or other enterprise databases.
Experience with Docker and cloud or deployment platforms such as AWS, Azure, GCP, Render, or Vercel.
Familiarity with CI/CD, automated testing, and production engineering practices.
Familiarity with open-source software development and collaborative engineering practices.

Key Skills

What We Offer

Opportunity to build production AI agents for enterprise data and analytics.
Hands-on experience with LLMs, agentic systems, MCP, and AI-powered analytics.
Exposure to natural-language-to-SQL, agent evaluation, tool orchestration, and production AI engineering.
Opportunity to contribute to TridaPad and AgentSQL and solve practical enterprise data problems.
A collaborative environment focused on building reliable and useful AI products.

Why Join Trida Labs?

At Trida Labs, you will work on AI systems that interact directly with real enterprise data and applications.

You will be involved across the agent engineering lifecycle — from agent architecture and tool development to NL-to-SQL accuracy, evaluation, optimization, and production deployment.

Equal Opportunity

Trida Labs is an equal opportunity employer. We are committed to creating an inclusive workplace where diverse perspectives are valued and everyone has the opportunity to contribute and grow.

Trida Network

AI Agent Engineer