Skip to content
AI Engineer · AI Automation Architect · London, UK

I build production AI that does real work.

From internal tools to production AI systems, built end to end.

Gbolagade Ishola

About

I'm Gbolagade, an AI engineer based in London. I build the software businesses actually run on, internal tools, automations, customer-facing products, and production AI systems, and I use AI to make them faster, cheaper, and smarter.

I am not boxed into one kind of work. Some days that is a RAG pipeline or a multi-agent automation that takes real actions. Other days it is an internal tool, a back-office integration, or an open-source library that solves a problem cleanly. Lately it has meant joining a company's internal apps into one assistant with a memory layer behind it, and running open-weight models locally to see how far a small model goes. I go where the business problem is, and I use the right tool for it.

I work end to end, from research and prototyping through to deployment, observability, and real business impact, across customer operations, internal tooling, ecommerce, and content. I care about systems that are accurate, cost-aware, and accountable, whether they are powered by a large language model or just good engineering.

What I Build & The Impact

Business software & internal tools

The day-to-day systems a business runs on: internal tools, dashboards, integrations, and automations, in custom code or on Power Platform, that remove manual work and connect the systems you already use.

Production LLM & agent systems

RAG pipelines, multi-agent orchestration, agent memory, MCP servers, and AI co-pilots that take real actions behind confirmation. Evals and structured-output validation keep them predictable in production.

AI for real business problems

Customer operations, content, ecommerce catalogs, and back-office workflows. I use AI where it pays off and plain engineering where it does not, then ship the result.

Observability, governance & open source

Latency, cost, and output-validity monitoring with GDPR-aligned redaction, plus open-source tools and local-model experiments I build and share. AI you can trust, audit, and run accountably.

Projects

Executive AI Assistant

Voice-and-chat AI co-pilot that triages email, runs the calendar, preps meetings, and places calls.

TypeScriptNext.jsOpenAI RealtimeMicrosoft Graph
View case study

Federated Agent Platform

A hub that joins a company's internal apps into one agent surface, with per-user permissions, audit, and a typed memory layer.

TypeScriptNext.jsPostgreSQLPrisma
View case study

Agentic Product Content Pipeline

Multi-agent ingestion, enrichment, and publishing with a Pinecone-backed RAG taxonomy.

PythonOpenAIPinecone (RAG)Docker
View case study

AI Customer Support Co-pilot

Self-hosted support stack with AI routing, drafted replies, and back-office automations.

PythonLLM APIsWebhooksDocker
View case study

Workplace Automation on the Microsoft Stack

Power Automate and Power Apps workflows, Copilot Studio agents, and desktop RPA for legacy processes.

Power AutomatePower AppsCopilot StudioCopilot Agent Builder
View case study

AI-Native CRM

Open-source CRM with AI co-pilots, automatic enrichment, and natural-language actions.

TypeScriptReactGraphQLPostgreSQL
View case study

Alice, AI Task and Content Studio

Tasks and multi-channel content in one studio, with a five-stage agent pipeline, a copilot, and an MCP server.

Next.jsTypeScriptVercel AI SDKSQLite
View case study

Local Agent Panel and Framework Benchmark

Researcher, Critic, and Chair deliberation on a local Qwen model, plus a tool-calling comparison of AgentScope and Qwen-Agent.

PythonAgentScopeQwen-AgentOllama
View case study

Multi-Provider LLM Proxy

Open-source FastAPI gateway with provider fallback, cost tracking, and structured logging.

PythonFastAPIOpenAIAnthropic
View case study

North Star Support Bot

Deterministic support agent in a demo storefront: order tracking, returns, recommendations, and human handoff, with no runtime dependencies.

TypeScriptViteVitestSingle-file build
View case study

AI Operations Monitor

Open-source FastAPI, Postgres, and Grafana stack for AI latency, cost, and failure tracking.

PythonFastAPIPostgresGrafana
View case study

Connect

Building something, or thinking about it? Book a 1:1 and let's dig in.

Gbolagade Ishola