I build AI automations for small teams: lead capture, email triage, support bots and custom agents. Fixed price. Delivered in 3–5 days. Zero calls. Watch the real thing run, live on this page.
Real output: a local open-source LLM (Qwen2.5-1.5B-Instruct via Ollama) classified each of 6 sample emails and wrote the reply drafts, unedited. Drafts are saved for review, never auto-sent. Client builds plug the same step into Gmail/Outlook through n8n with OpenAI, Claude or a self-hosted model.
Real output: retrieval-augmented generation. Local embeddings (nomic-embed-text) pick the best matching doc out of 5 sample docs, and a local LLM (Qwen2.5-1.5B) writes the answer. If the match is weak, the bot hands off to a human instead of guessing.
Live scrape of books.toscrape.com (a public practice site), 60 rows from 3 pages, first 8 shown. The summary is computed from the data. Client builds can push to Google Sheets or Airtable on a schedule.
Pick a build on Contra and describe the process you want gone.
I reply in writing with exactly what gets built. Price stays fixed.
3–5 days. You get progress updates and a short screen recording.
You own the workflow, plus written docs for running and changing it.