Building products with AI as a force multiplier
I use AI to build better products, faster.
I combine full-stack engineering, product thinking, and AI-assisted workflows to research, prototype, build, and ship thoughtful software products.
AI-assisted product workflow
shipping
My product-building workflow
AI supports the work without replacing the product judgment, design taste, or engineering ownership.
1Idea
2Research
3Define
4Prototype
5Design
6Build
7Ship
8Iterate
AI enhancesworkflow
✓Research
✓Product Decisions
✓Project Context
✓Knowledge
✓Vibe Coding
✓Testing
✓Code Review
✓Documentation
Featured project
A RAG system powering an interactive portfolio.
A full-stack retrieval system that turns project notes, technical writing, and site content into a conversational experience.
RAG system case study
live on siteProblem
Static portfolios hide context across pages, repos, notes, and posts.
AI role
Retrieval and generation work together to answer questions from my own knowledge base.
Product UX
You can explore my projects, process, and technical decisions through conversation.
Shipped system
FastAPI RAG backend, embedded chat widget, full chat page, streaming responses, and conversation memory.
FastAPINext.jsLlamaIndexLangChainOpenAIDeepSeekHybrid searchPostgreSQLPineconeLangSmithRAGASQuery rewritingRerankingStreaming responsesIncremental ingestionMulti-source loadersConversation context
Roger's AI, powered by RAG
I built a retrieval-augmented system that connects my portfolio, project notes, and technical writing to a conversational version of me. Behind the interface, it handles multi-source ingestion, retrieval, reranking, conversation context, streaming responses, LangSmith tracing, and RAGAS-backed evaluation.
Hybrid retrieval
Conversation memory
Incremental ingestion
Evaluated source-grounded responses
Current focus
What I am sharpening across product and engineering.
I am focused on building better product instincts, stronger software systems, and practical AI workflows that help ideas become reliable shipped products.
Product discovery
Using AI to turn ambiguous ideas, user needs, notes, and constraints into clearer product direction.
Rapid prototyping
Moving from concept to working interaction faster with AI-assisted exploration, UI iteration, and implementation.
Engineering workflows
Leveraging AI to improve efficiency and quality across coding, code review, testing, documentation, automation, and delivery loops.
Systems foundations
Strengthening architecture, databases, cloud, security, distributed systems, and AI fundamentals for reliable delivery.
Selected work
See how I turn ideas into usable products.
Browse a few projects that show how I think through product direction, interface design, engineering tradeoffs, and shipped workflows.
Journal
Read how I think through product and engineering.
Build notes on AI-assisted workflows, system design, UX decisions, and the tradeoffs behind shipping software products.
Sep 22, 2026
AIAgentHarnessArchitecture
From One Agent to an Enterprise SaaS Agent - How It Becomes Resilient and Scalable
> Managed Agents are easier to understand by starting with a local Agent. Once it needs to serve more users, run for longer, and recover reliably from component failures, its execution loop, durable record, and execution environment need to be separated so that each can be recovered and replaced ind
Sep 15, 2026
AIHarnessAgent
Understanding Agent Harness Step by Step
> An agent harness is the software around an LLM that runs the execution loop, executes tool calls, and manages the context and state needed to continue a task. The model decides what to do next; the harness coordinates how that decision is carried out. To understand why these pieces are needed, I f
Sep 10, 2026
AIAgentTool
Notes on Writing Effective Tools for AI Agents
A tool is an external capability an AI agent can call to perform actions or access information beyond what the model can do by itself. Designing an effective tool means making it clear when to use it, how to call it, and what the agent can learn from its response. LLM-oriented design In traditional
Work together
Building a product that should move faster with AI?
I am interested in product builds where AI helps with research, automation, knowledge, UX, and delivery, as well as AI product experiences where intelligence is part of the core interface.