QwkSearch
Production RAG systems that turn large document libraries into searchable, grounded answers and useful research workflows.
Full-stack AI developer / Bay Area & remote
Functionally brilliant.Elegantly simple.
I build end-to-end software across AI, web, and data—from document ingestion and agent tooling to polished interfaces and dependable cloud delivery.
10+
years building software
50+
tools and packages
500K
monthly downloads
6K+
documents processed daily
Selected work
Tools for the messy middle.
A few ways I turn hard problems into clear, useful products—from AI infrastructure to open-source tools and client work.
DisasterAid.ai
A climate-research RAG assistant built with Transformers.js and SvelteKit. Awarded the AWS Climate Tech Top Prize at the Berkeley AI Hackathon.
Debate-AI.com
A crowdsourced argument-mapping platform for students, pairing structured debate with LLM-assisted scoring.
Rights Institute / 1computer.online
A maintained collection of more than 50 practical web tools and NPM packages, reaching 500K monthly downloads.
Welcome Home Agency
Web development for a real-estate client platform, bringing property discovery and a welcoming brand experience together.
The work behind the work
Built in the real world.
A decade of shipping across early-stage AI, climate tech, autonomous vehicles, and enterprise data products.
Aug 2025—2026AI DeveloperQwkSearch · San Francisco, CA / RemoteDocument intelligence · AI agent infrastructure
- Built OCR and structured-extraction pipelines processing 6,000+ documents per day at approximately 96% accuracy, using PDF parsing, layout detection, and transformer-based OCR.
- Combined chunking, embeddings, vector search, reranking, caching, and grounding controls into RAG systems; increased document-processing throughput by 30%.
- Developed MCP-compatible research workflows, repository summarization, and evaluation for retrieval quality and answer grounding.
- Deployed containerized AI services and automated delivery across Cloudflare, AWS, GCP, Terraform, and GitHub Actions.
Jan—Aug 2025AI DeveloperBioGas Engineering · Los Angeles, CA / RemoteClimate intelligence · Emissions analysis
- Built a SvelteKit RAG chat application with Transformers.js and custom scrapers, reducing research query time by 40%.
- Applied Python machine-learning models to forecast biogas-system emissions and improve environmental compliance workflows.
- Connected extraction, vector retrieval, predictive modeling, and an accessible frontend into one operational platform.
2023—2024AI DeveloperTrieve.AI · San Francisco, CA / Remote · YC W24AI agent infrastructure
- Built an MCP client for connecting external services and added file summarization to agent workflows working across large repositories.
- Designed runtime-aware routing across vector databases, making indexing and search more portable across environments.
- Improved indexing throughput with worker-based parallelism and memory controls, eliminating out-of-memory failures.
2020—2022AI DeveloperOttometric · San Francisco, CA / RemoteAutonomous-vehicle data tooling
- Built a full-stack error-detection system for lidar and camera data, including its server infrastructure and user interface.
- Created an interactive WebGL canvas to inspect object- and lane-detection data with rotation and panning controls.
- Deployed computer-vision prototypes behind AWS EC2 REST APIs and used CloudWatch to monitor production endpoints.
2015—2019AI DeveloperPro Unlimited · San Francisco, CA / RemoteCorporate HR platform · Data engineering
- Automated competitor-pricing research with Selenium and BeautifulSoup, custom Python connectors, and structured data pipelines.
- Moved Excel workflows into Python machine-learning pipelines, reducing runtime by 60%; shipped Flask dashboards and data visualizations.
- Built feature-engineering and model-comparison workflows, and worked with QA to raise ML module test coverage to 85%.
Open a role to see selected highlights.
Notes from the work
Ideas worth following.
Research on collective intelligence, agentic systems, decision-making, and how the web can help people think together.
- 01Is AI a Conscious Being With Rights?: Emergence of Post-Human Collective Consciousness.
- 02QwkSearch.com Reimagines the Web as a Self-Organizing Mind Map.
- 03Debate AI Collective Consciousness: Reimagining the Web as Argument Mind Map Is Key to Crowdsourced AI Decision-Making.
- 04Automated Investment Broker: Research Agents Debate, Correlate, and Copy Prediction Market Elite Forecasters.
- 05The Synergy Slayer of Debate Critical Theory: Putting Theory Into Action Avoids Distraction by Abstraction Justifying Inaction.
A little context
A builder with a research habit.
From data pipelines and interfaces to AI infrastructure, I like working where the product is still taking shape.
I'm Alex, a full-stack AI developer with more than a decade of experience across Bay Area startups and product teams. I bring market research, feature design, and engineering together to make tools that are functionally brilliant and elegantly simple.
My work spans the full product loop: turning messy documents into grounded answers, connecting agents to useful tools, building intuitive web and mobile experiences, then deploying and improving them in production.
The toolkit
Tools for the whole stack.
AI, agents & retrieval
- Agentic workflows
- MCP
- RAG architecture
- Document ingestion
- Semantic & vector search
- Embeddings
- Evaluation & grounding
- LangChain
- Transformers.js
- OpenAI API
- Python ML
- FBProphet
- Claude
- Perplexity
- Cursor
- Antigravity
- v0
Product engineering
- TypeScript
- JavaScript
- React
- Next.js
- SvelteKit
- Node.js
- Bun
- Python
- React Native
- Expo
- HTML & CSS
- Tailwind UI
- WebGL
- Chrome extensions
- OpenAPI & REST
- Google APIs
Cloud, data & delivery
- Cloudflare
- AWS
- GCP
- Vercel
- Docker
- PostgreSQL
- Supabase
- GitHub Actions
- nginx
- Caddy
- Shopify
- Autoscaling
Have a good problem?
Let's code the thing already.
For thoughtful teams, ambitious ideas, and systems that deserve to feel simpler.