You will get Build a Python data workflow and deliver clean CSV or JSON
Rising Talent
Project details
You will get a clean, structured dataset or reusable Python data workflow delivered in CSV or JSON and ready to use. I focus on small, fast-turnaround data tasks where you need cleanup, normalization, field mapping, deduping, and a reliable final export without turning it into a large engagement. This is a good fit when you already have source material, exports, listings, or public data that needs to be cleaned and organized. I can deliver the finished files only or include a lightweight Python workflow so the process is easy to rerun.
Programming Languages
PythonWhat's included
| Service Tiers |
Starter
$100
|
Standard
$275
|
Advanced
$675
|
|---|---|---|---|
| Delivery Time | 2 days | 3 days | 5 days |
Number of Revisions | 1 | 1 | 2 |
Number of Pages Mined/Scraped | 1 | 3 | 10 |
Number of Sources Mined/Scraped | 1 | 2 | 4 |
Install Script | - | - | |
Test Script | - | - | |
Task Automation | - |
About Dylan
AI Agent Developer | Claude & OpenAI, MCP, RAG & Automation
Brooklyn, United States - 5:30 pm local time
I specialize in agents specifically, and I work with the newest agent tooling every day: the Claude Agent SDK, the OpenAI and Vercel AI SDKs, and the Model Context Protocol (MCP) for connecting agents to your own systems.
What I can build for you:
- Custom AI agents that use tools, call your APIs, and complete tasks end to end
- MCP servers that securely connect an agent to your internal tools, databases, and SaaS (Slack, Notion, CRMs, custom APIs)
- Document and knowledge assistants (RAG) that answer questions over your contracts, docs, and data
- Workflow automation: intake and triage, research, data extraction, report and email drafting, internal copilots
- The production layer most freelancers skip: evals, guardrails, logging and monitoring, and error handling — so the agent is reliable, not a one-off demo
My stack:
- AI / Agents: Claude (Agent SDK), OpenAI, Vercel AI SDK, MCP, tool-use, RAG, evals and guardrails
- Backend: TypeScript/Node, Python, Express, REST APIs, PostgreSQL, MongoDB, Supabase
- Infra and ops: Docker, CI/CD (Vercel, Render), observability (Grafana/Loki, Sentry), Linux
An edge that helps on sensitive projects: before moving into engineering I worked in legal and risk/investigations (Paul, Weiss and Kroll), so I understand document-heavy, detail-critical, compliance-aware workflows — and I build agents that respect them.
How I work:
1. A short call to map the exact workflow you want to automate
2. A scoped plan with clear milestones — and an early working demo so you see progress fast
3. Build, test against real evals, deploy, and hand off with documentation your team can maintain
If you have a workflow you think an AI agent could handle — or you're not sure whether it can — message me a few sentences about it and I'll tell you honestly what's possible and how I'd approach it.
Steps for completing your project
After purchasing the project, send requirements so Dylan can start the project.
Delivery time starts when Dylan receives requirements from you.
Dylan works on your project following the steps below.
Revisions may occur after the delivery date.
Confirm inputs and output schema
I review the source material, confirm the exact fields and output structure, and align on a clean sample before building the final workflow.
Clean and structure data
I process the source material, normalize fields, handle duplicates or format issues, and structure the data for a reliable final output.