Eiman Abdelmoneim
Solutions Architect & Ai Lead (Growth & Gtm) @Thirdweb
Signup · Get unlimited contacts
WORK HISTORY
Solutions Architect & Ai Lead (Growth & Gtm) @Thirdweb
Leading AI Growth & GTM Engines (\"Billie\" & \"Siwa\"): Architected the company\'s internal AI workforce, integrating LLMs directly with our core Enterprise SaaS stack to drive revenue and support automation. • Billie (RevOps Agent): Built the outbound engine that orchestrates actions across HubSpot, Google Workspace, and Slack. Designed deterministic \"Tool Use\" definitions (Function Calling) to allow the agent to safely read CRM deals and draft emails without hallucinating data. • Siwa (Support Agent): Engineered an \"always-on\" support agent handling 60k+ conversations. Built a federated RAG pipeline that ingests data from Notion (Engineering Wikis), Intercom, and Linear to autonomously resolve queries and file tickets.• Architecting the \"Agent Engine\" for Prediction Markets: Currently leading the AI architecture for a new crypto trading platform. Building a deterministic \"Analyst Agent\" that synthesizes real-time on-chain data to provide traders with neutral, data-backed insights.• Infrastructure & Observability: Established the \"Agent-Led Reporting\" standard, where agents generate their own weekly ROI reports. Standardized the stack on LangGraph, Supabase (pgvector), and BullMQ.
EDUCATION
University of Illinois Chicago
MS, Computer Science
University of Illinois Urbana-Champaign
BA, Math/CS
SKILLS
ABOUT EIMAN ABDELMONEIM
I am an Applied AI Architect who sits at the intersection of Product Strategy and Engineering.I don\'t just build demos; I architect and ship end-to-end GenAI systems that survive production. My focus is on moving beyond the hype to build \"Pragmatic AI\"—systems that anchor probabilistic LLMs to deterministic Enterprise workflows, ensuring they are safe, scalable, and drive measurable business value.Currently, I operate across two high-impact domains: At thirdweb: Leading Enterprise AI & Growth EnginesI lead the architecture and deployment of the company\'s AI workforce, integrating LLMs directly with our Enterprise SaaS stack (HubSpot, Linear, Notion):• The Growth Engines (Billie & Siwa): I architected the autonomous agents that run our Revenue Operations and Customer Success.\"Billie\"(RevOps) orchestrates complex workflows across HubSpot and Gmail, using a \"Supervisor Model\" to score draft quality.\"Siwa\"(Support) handles 60k+ conversations, using a federated RAG pipeline to resolve queries and autonomously file tickets in Linear.• The Product Engine (Crypto Prediction Markets): I am currently building the Agent Engine for a new prediction market platform. This is a high-stakes, quantitative environment where I build deterministic \"Analyst Agents\" that ingest real-time on-chain data to provide neutral, actionable market insights.🧠 At humansinloop.xyz: Founder & Lead ArchitectI use this platform as my Applied AI R&D Lab to solve the \"soft\" problems of AI—mentorship, tone, and empathy.• Long-Term Memory Architecture: Engineered a persistent memory system using LangGraph and Supabase (pgvector). Built an asynchronous \"Observer\" pattern that analyzes conversations to index key user facts (goals, skills), enabling hyper-personalized, multi-session mentorship without context loss.• The \"Data Flywheel\" Methodology: Pioneered a \"Hall of Fame\" workflow to curate golden datasets from best-in-class interactions, using them for Few-Shot Prompting and Trajectory Evals.My Engineering Philosophy:• Pragmatism over Magic: Focus on \"low-hanging fruit\" where humans act like LLMs.• Deployment is Data Creation: Every agent I build has a mature data lifecycle—evolving from Zero-Shot to Fine-Tuned models.• Agent-Led Observability: I don\'t just monitor latency; I build agents that report on their own ROI (e.g,\"I resolved 50 tickets this week\").The Stack:LangGraph, LangChain, TypeScript/Node.js, Python, Supabase (pgvector), OpenAI/Anthropic SDKs, Model Context Protocol (MCP)
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.