Certifiedauthur

AI Engineer

Kuuka • Freelance

Nov 2025 - Feb 2026 • 3 mos

• Built and deployed multiple AI-driven automation systems (supplier intelligence, order optimization, and market monitoring) that enabled data-driven purchasing decisions, automated supplier evaluation, and surfaced cost-saving opportunities across procurement workflows. • Architected and implemented a multi-channel notification and campaign platform supporting segmentation, analytics, templates, and automated delivery pipelines, enabling marketing and operations teams to run targeted campaigns and track engagement across tenants. • Designed and integrated enterprise-grade integrations and AI tooling (Xero, GA4, Calendly, WhatsApp automation, and MCP tools), expanding product capabilities and enabling autonomous agents to interact with financial systems, analytics, and external services. • Improved system reliability and deployment stability by hardening Docker builds, fixing service orchestration issues, implementing CI-ready tests across ~50+ endpoints, and resolving runtime dependency failures that previously blocked deployments. • Implemented security, compliance, and observability infrastructure including audit logs, incident management, GDPR export flows, PII redaction, and RBAC-based tenant isolation, strengthening platform governance and regulatory readiness.

AI ENGINEER

Zenovo-AI • Freelance

Oct 2024 - Oct 2025 • 1 yr

• Designed and implemented an AI-driven proposal generation platform (FastAPI + LangChain/LangGraph + OpenAI) with multi agent workflows (intent routing, RAG retrieval, structural drafting, reflexion/critic loop) to automate RFQ/RFP response creation. • Built scalable retrieval architecture using PostgreSQL + pgvector and optimized document ingestion (PDF parsing, Google Drive/Docs integration, semantic chunking) to enable high‑relevance, low‑latency context retrieval for generation. • Engineered performance layer with Redis caching, Celery background/periodic workers, and database connection pooling; cut repeated query and generation latency dramatically while supporting asynchronous long‑running tasks. • Delivered production-grade DevOps: Dockerized micro-services, blue‑green deployments via Nginx, and a GitHub Actions CI/CD pipeline (tests, security scanning, load testing, automated rollback) enabling zero‑downtime releases and rapid iteration. • Enforced robustness and observability with Pydantic‑validated LLM outputs, health & metrics endpoints, cache/stat introspection APIs, structured logging, and database optimization routines (index & full‑text strategy) to sustain growth and maintain reliability.

AI Developer

Junify • Freelance

May 2025 - Aug 2025 • 3 mos

• Architected a multi‑agent conversational AI platform using FastAPI, LangChain, and LangGraph; implementing LLM-driven intent routing, context summarization, and dynamic workflow branching to reduce irrelevant agent calls and improve response relevance. • Engineered automated chart generation and data retrieval pipeline (SQL + vector similarity search) that transforms natural language queries into structured insights and frontend-ready visualization configs, accelerating analytical turnaround time. • Designed and enforced robust state management with checkpointed graph execution (PostgreSQL) and background memory summarization, enabling persistent conversational intelligence and lowering token usage via strategic history compression. • Implemented human-in-the-loop decision agents (table selection + interrupt handling) and context sufficiency evaluation, increasing accuracy of downstream retrieval and reducing unnecessary database scans across multi-tenant datasets. • Refactored and stabilized asynchronous API endpoints with resilient error handling, type-safe state propagation, and structured JSON outputs (including chart configs), improving system reliability and integration readiness for production deployment.