<div>Welcome to Zealant Consulting Group, your trusted partner in recruitment solutions. At ZCG, we believe that the heart of every successful organization is its people. With a passion for connecting talent with opportunity, we have established ourselves as a leading recruitment business dedicated to fostering growth and success for both individuals and businesses. We aim to inspire, shape, and rethink the way leading brands does recruitment to grow their business. We are a team of young professionals, passionate to solve complex recruitment problems. It’s about identifying, matching, mapping, and locating the most-suitable leader to captain the organizational roadmap of businesses. At ZCG, we’re the compass to this synergy, guiding the global leadership hiring landscape.</div>

Zealant Consulting Group

Job Description: Agentic Workflows Engineer Role Overview
You will design, build, and scale the agentic workflows that power client’s CLM intelligence platform. This is an engineering role for someone who thinks in systems: multi-agent orchestration, tool use, state management, guardrails, and failure recovery. You’ll work directly with Product and Applied AI to turn complex contract workflows (intake routing, clause extraction, obligation monitoring, negotiation assist) into reliable, production-grade agentic pipelines.
Key Responsibilities
<ul>
<li>Design and implement multi-agent pipelines using frameworks like Lang Graph, Crew AI, AutoGen, or equivalent—taking systems from design through production deployment.</li>
<li>Build tool-use and function-calling architectures that connect agents to CRM, ERP, CLM, and procurement systems in real time.</li>
<li>Implement human-in-the-loop (HITL) checkpoints, escalation logic, and guardrail layers that keep agents on-policy without breaking flow.</li>
<li>Build stateful agent memory, including short-term context management, long-term retrieval, and task persistence across sessions.</li>
<li>Develop observability and tracing so the team can understand when an agent made the right call (and when it didn’t).</li>
<li>Create evaluation harnesses to measure agent behavior across edge cases—not just happy paths.</li>
</ul>
Qualifications Required
<ul>
<li>3+ years of software engineering experience, with 1+ year working on LLM-based systems in production.</li>
<li>Hands-on experience building agentic or multi-agent systems (not just chatbots or RAG pipelines).</li>
<li>Strong Python skills; comfortable with async patterns, API design, and distributed systems fundamentals.</li>
<li>Solid understanding of LLM behavior in production: context windows, tool calling, structured output, and prompt stability under load.</li>
<li>Experience with at least one agentic framework: LangChain/LangGraph, AutoGen, CrewAI, Semantic Kernel, or similar.</li>
<li>Ability to reason about failure modes (e.g., agent loops, tool misuse, context drift, confident incorrect outputs) and design recovery paths.</li>
</ul>
Preferred
<ul>
<li>Experience in document-heavy domains.</li>
<li>Familiarity with vector databases (e.g., Pinecone, Weaviate, pgvector) and hybrid retrieval approaches.</li>
<li>Background with workflow orchestration tools (e.g., Temporal, Airflow, or similar).</li>
<li>Experience deploying agents that operate across third-party SaaS integrations.</li>
</ul>
Working Style
<ul>
<li>Small team, high autonomy—you own your pipelines end to end.</li>
<li>Decisions are made with working code, not slide decks.</li>
<li>You’ll see your work used by procurement and legal teams at large enterprises within weeks of shipping.</li>
</ul>
Success Metrics
You’ve shipped an agentic system that delivers real value in production (beyond demos or hackathon projects). You understand where agents break down (e.g., tool misuse, context drift, hallucinated decisions) because you’ve diagnosed and fixed those issues. You care about reliability and measurable outcomes as much as capability.
&nbsp;