Marketing / Sales AI agent Builder

navneetkaur | Updated: June 16, 2026

About Trantor:

Trantor is a technology services company focused on outsourced product development and digital re-engineering. Leveraging our CaptiveCoE™ engagement model, we operate as a seamless extension of our client’s teams to provide rapid scalability with predictable budgets. Founded in 2012, Trantor has worked with customers across Tech, FinTech, Media & Cybersecurity industries. We have centers in the US, India, Canada, and Costa Rica. We are consistently rated as the #1 employer in the region with the ability to attract and retain technical talent. Our commitment to excellence and impactful results has translated to long-term relationships and value for our clients and solution partners. 

 

Job Overview

As our Marketing AI Agent Builder, you will be the lead architect for our agentic marketing stack. Your goal is to move beyond simple automation (if-this-then-that) toward autonomous systems that reason through marketing data, execute multi-channel campaigns, and self-correct based on performance. This is a "Builder"
role—part Engineer, part Strategist, part Data Scientist.

 

Core Responsibilities

  • Intelligent Research & Segmentation: Deploy agents to crawl the web for intent signals (hiring changes, funding, news) to dynamically update ICP segments and enrich CRM records in real-time.
  • Autonomous Campaign Management: Build agents that independently monitor live campaigns (Search, Social, Email) and autonomously reallocate budgets or pause underperforming creatives based on real-time ROI.
    Dynamic Lead Nurturing: Design agents that use Natural Language Processing (NLP) to “read” prospect behavior and trigger personalized, multi-step nurture sequences without manual intervention.
    Performance & Guardrails: Establish “human-in-the-loop” protocols and evaluation frameworks to ensure agent outputs align with brand voice and security standards.
    System Orchestration: Connect AI agents to our core stack—including HubSpot/Salesforce, 6sense, and Google Analytics—ensuring seamless data flow and automated write-backs to the CRM.
 
Qualifications &Skills

Role Overview:

You will architect and manage a fleet of autonomous Sales Agents that perform end-to-end prospecting, qualification, and meeting scheduling. Unlike traditional Sales Ops, you won’t just build workflows; you will build reasoning systems that research prospects, handle objections, and manage their own "inboxes" to deliver qualified meetings to our SEs.

Key Responsibilities

  • Autonomous Lead Qualification: Design agents to “vet” inbound leads by querying external data sources and internal CRM history to determine if they meet our Ideal Customer Profile (ICP).
    Inbox Management & Objection Handling: Deploy agents (using LLMs like GPT-4o or Claude) that can read incoming emails, categorize intent, and draft/send context-aware replies to handle common objections.
    Live Sales Assistant (RAG): Build a “Sales Knowledge Base” using Retrieval-Augmented Generation (RAG) so agents can instantly reference case studies, pricing, and competitor battlecards during autonomous chats.
    CRM Data Integrity: Ensure agents “write back” all interactions to Salesforce/HubSpot, keeping data clean and attribution accurate without rep intervention.

    Qualifications
    • Experience: 4+ years in Growth or Solutions Engineering in a B2B SaaS environment.
    • AI Tooling: Experience with “Agent-as-a-Service” platforms (e.g., 11x, Regie.ai, Lyzr) or building custom agents via LangChain / LangGraph.
    • Sales Stack Mastery: Deep expertise in Salesforce/HubSpot, Salesloft/Outreach, and data providers like ZoomInfo or 6sense.
    • Analytical Rigor: Ability to A/B test agent prompts and “reasoning loops” to improve meeting-set rates.
     
     
Job Category: Claude CrewAI GPT-4o HubSpot LangChain LangGraph RAG (Retrieval-Augmented Generation) Salesforce Zapier
Job Type: Full Time
Job Location: Remote
Shift Timing: 8.00PM - 4.00 AM PST

Apply for this position

Allowed Type(s): .pdf