Agentic AI Engineer

Lumovy Technology Solutions
Lahore
Workplace: OnsiteFull timeFunction: Data Science & Machine LearningSkills: ["Clear written communication","Test-focused mindset","Technical design to implementation","Collaboration","Asking clarifying questions"]

Build enterprise AI agentic solutions on a code-first agent runtime, including the multi-stage decision loop and governed tool layer for reading and carefully writing to systems of record. Implement grounding-focused prompts, human-approval gates, and resilient integration behaviors. Own features end-to-end—from design through implementation, tests, and agent evals—while tuning cost and latency for reliable, explainable outcomes in production.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Lumovy Technology Solutions
Lumovy Technology Solutions
4 hours ago

Agentic AI Engineer

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 4 hours agoStatus: Live

Job Summary

Build enterprise AI agentic solutions on a code-first agent runtime, including the multi-stage decision loop and governed tool layer for reading and carefully writing to systems of record. Implement grounding-focused prompts, human-approval gates, and resilient integration behaviors. Own features end-to-end—from design through implementation, tests, and agent evals—while tuning cost and latency for reliable, explainable outcomes in production.
Location: Lahore
Workplace: Onsite
Employment Type: Full time
Job Function: Data Science & Machine Learning
Seniority: Mid level

Key Responsibilities

  • •Implement agent behavior on a code-first Python runtime, including the multi-stage decision loop that observes, explains, recommends, approves, executes, measures, and learns.
  • •Design and write grounded prompts so the agent cites live evidence from real data and avoids fabricated outputs.
  • •Implement the human-approval gate so no connected system is modified without explicit, auditable user confirmation.
  • •Build and extend the governed tool layer (via Model Context Protocol) to connect agents to business systems using scoped read tools and gated write actions under delegated identity.
  • •Create unit tests and agent evals for task success, grounding/faithfulness, approval integrity, and cost/latency against defined SLOs.

Key Requirements

  • •3–5 years building production software in Python with a track record of shipping and maintaining real systems.
  • •Hands-on experience building LLM-powered features or agents, including LLM APIs, tool/function calling, and reliable agent loops.
  • •Comfort integrating with external systems and handling auth via OAuth/OIDC-style delegated flows, including messy upstream responses and partial failure.
  • •Strong testing instincts, including unit tests and the ability to test non-deterministic agent behavior.
  • •Clear written communication to turn technical designs into reviewed code and explain trade-offs in documentation and code review.
Experience:Enterprise softwareLLMAI agentsAgentic AI
Skills:Clear written communicationTest-focused mindsetTechnical design to implementationCollaborationAsking clarifying questions
Tech Stack:PythonLLM APITool/function callingPromptsModel Context ProtocolMCPOAuthOIDCAzureAWSGCPModel Context Protocol (MCP)Dynamics 365SAPNetSuiteSalesforceMicrosoft TeamsSlack

Company Brief

Lumovy Technology Solutions
Provides custom software development, cloud and IT consulting services focused on digital transformation, application development, and managed technology solutions for businesses seeking scalable, modern technology and integration support.
Industry: Professional Services
Website