Forward Deployed Engineer - Retail

Forward Deployed Engineer - Retail (AI Architect - Agentic AI & Digital Platforms)

Dallas, TX (Preferred)

Onsite | 12–18 Years Experience

Overview:

Founded in 1986, Mastech is a NYSE American company (MHH) that helps leading organizations become Knowledge-First enterprises. We help businesses harness the knowledge of their people, business, and systems to create measurable value with AI. By building trusted data foundations, applying AI to real business challenges, and combining deep expertise with the right technology, we help organizations make better decisions and achieve meaningful outcomes.

Our work spans healthcare and life sciences, banking and financial services, consumer and retail, energy and utilities, and manufacturing and logistics. With a global team serving more than 150 enterprise customers, Mastech brings together AI, data engineering, cloud technologies, and enterprise platforms to help people and businesses Thrive on What’s Next.

Website

http://www.mastech.com

Role Overview

A hands-on techno-functional architecture role - not research, not purely strategic. The Senior Principal Architect serves as the onshore technical anchor for an Agentic AI initiative across customer-facing digital platforms: translating business intent into scalable, governed, production-ready systems and holding the architectural line across delivery teams. The right person codes, reviews, and debates trade-offs in the same week they brief a VP.

Key Responsibilities

• Design and govern end-to-end agentic architecture for digital experiences spanning apps, loyalty, payments, and commerce workflows.

• Architect multi-agent orchestration patterns - orchestrator/subagent hierarchies, tool/function calling, stateful graph-based workflows, and agent memory (short-term, session, long-term).

• Establish MCP (Model Context Protocol) standards for governed tool and enterprise service exposure.

• Define runtime guardrails: spend controls, fraud detection hooks, human-in-the-loop checkpoints, and auditability trails.

• Lead AI agent integration with retail systems - catalog, inventory, pricing, promotions, loyalty, and checkout.

• Set LLMOps standards: prompt versioning, RAG pipeline governance, evaluation harnesses, and observability (latency, token cost, drift, safety).

• Serve as the primary technical interface between business leaders, delivery teams, and platform partners.

Required Qualifications

Foundational

• 12–18 years in enterprise/platform architecture; 3–4+ years delivering AI/ML or GenAI systems in production.

• Strong command of API-first design, distributed systems, event-driven architecture, and real-time data.Expertise in building and scaling applications.

Agentic AI - Core Requirement

• Production experience with LLM-based agent systems - shipped, not just prototyped.

• Hands-on with agent frameworks: LangGraph, LangChain, Semantic Kernel, CrewAI, or AutoGen - and judgment on agentic architectural patterns

• Working knowledge of: ReAct/plan-execute patterns, multi-agent handoffs, RAG pipeline design (chunking, hybrid retrieval, reranking), context/prompt engineering, and MCP.

• Familiarity with agent failure modes - hallucination, tool misuse, cost runaway, infinite loops - and how to architect against them.

• LLMOps exposure: evaluation frameworks, observability tooling (LangSmith, Langfuse, or equivalent), prompt/model lifecycle management.

Retail / Consumer Domain

• Meaningful exposure to retail commerce systems: catalog, pricing, promotions, loyalty, checkout, or payments. Direct retail experience strongly preferred; adjacent commerce background considered.

Data & Integration

• Comfortable with event-driven architectures (Kafka or equivalent) and enterprise data quality realities.

• Familiarity with semantic models or knowledge graphs as agent grounding layers is a plus.

Leadership Profile

• Translates agentic AI concepts into business value language credibly at the executive level.

• Sets architectural boundaries, makes trade-off calls, and defends them under pressure.

• Experienced leading across onshore/offshore delivery models.

• Hands-on enough to write reference implementations and lead code reviews - Python proficiency expected.

• Comfortable operating in a fast-moving technology space with evolving standards.