I am an AI Engineer with 3 years of production experience building LLM-powered systems that serve thousands of daily users. I specialize in RAG pipelines, multi-agent architectures, and scalable APIs. I am a strong Python developer with hands-on experience in LangChain, FastAPI, and vector databases. I am eager to leverage my background in generative AI and system design to build impactful AI products.

Aniekan Udo

I am an AI Engineer with 3 years of production experience building LLM-powered systems that serve thousands of daily users. I specialize in RAG pipelines, multi-agent architectures, and scalable APIs. I am a strong Python developer with hands-on experience in LangChain, FastAPI, and vector databases. I am eager to leverage my background in generative AI and system design to build impactful AI products.

Available to hire

I am an AI Engineer with 3 years of production experience building LLM-powered systems that serve thousands of daily users. I specialize in RAG pipelines, multi-agent architectures, and scalable APIs.

I am a strong Python developer with hands-on experience in LangChain, FastAPI, and vector databases. I am eager to leverage my background in generative AI and system design to build impactful AI products.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate

Work Experience

AI Engineer at Loreon Technologies
November 1, 2022 - December 1, 2025
Built enterprise RAG system serving 2,000+ daily queries across 15,000+ documents with 97% answer accuracy using hybrid search; developed multi-agent workflow automation with LangGraph reducing document processing time by 60%; deployed FastAPI services with 100+ concurrent users, latency <200ms, and 99.5% uptime; reduced AI hallucinations by 40% via re-ranking, grounding and confidence scoring; engineered multi-tenant SaaS with isolated vector namespaces and RBAC; built LLM evaluation framework in CI/CD automating 50+ test cases; designed ETL for MongoDB to PostgreSQL for analytics dashboards; implemented auto-ingestion pipelines with PGVector for embeddings and automatic reindexing.

Education

Bachelor of Science at University of Port Harcourt
January 1, 2014 - January 1, 2019
Bachelor of Science at University of Port Harcourt
January 1, 2014 - January 1, 2019

Qualifications

AI/ML Professional Program
January 1, 2024 - February 7, 2026
LangChain for LLM Development
January 1, 2023 - February 7, 2026
Building Systems with ChatGPT API
January 1, 2023 - February 7, 2026
Applied Data Science Lab
January 1, 2023 - February 7, 2026

Industry Experience

Software & Internet
    BrandGuard AI: Marketing Content Creator Bot

    BrandGuard AI is a production-ready, multi-agent marketing content generation system designed to create brand-aligned content that improves over time through human feedback. It leverages RAG (Retrieval-Augmented Generation) to ground generation in your specific brand voice and a self-learning loop to adapt to your style.

    Key Features:

    • Multi-Agent Architecture: Powered by CrewAI, utilizing specialized agents for research, writing, and review.

    • RAG for Brand Alignment: Integrated with LlamaIndex to retrieve context from uploaded brand documents (PDF, DOCX, TXT).

    • Self-Learning Feedback Loop: Automatically adapts to brand style based on human approvals and feedback.

    • Production Resilience:

      • Circuit Breakers: Protects against cascading failures in external APIs (Groq, Tavily).
      • Rate Limiting: Integrated token-bucket limits for API protection.
      • Retry Logic: Exponential backoff for database and API operations.
      • Persistent Storage: All brand documents are stored in PostgreSQL for persistence across deployments.
      • Real-time Metrics: Analyzes brand tone, sentence structure, and signature phrases.

    Tech Stack

    • Framework: FastAPI (Python)
    • AI/LLM: Groq (Llama-3), CrewAI, LlamaIndex
    • Database: PostgreSQL (SQLAlchemy + psycopg2)
    • Monitoring: Structlog (Structured Logging)
    • Deployment: Railway

    Link to project:
    https://www.twine.net/signin