Scan the documents
Bring the transaction’s documents into the workflow as scans.
I’m Sid, a workflow builder based in Ottawa. I connect documents, AI tools and everyday business tasks into practical workflows, from reading information to uploading, submitting and printing.
In a paperwork-heavy government-sector role, I built a workflow I used in my own work for transactions that needed completed forms and a paper record. It connects document scanning, local OCR, Copilot with Jev (TypeSafe AI) for form matching, and printing after human confirmation. In my own use, it roughly halved transaction processing time and improved accuracy.
Bring the transaction’s documents into the workflow as scans.
Run OCR locally to extract the document text for the next step.
Use Copilot with Jev (TypeSafe AI) and stored forms to match the paperwork to the task.
Keep a human confirmation step before the completed paperwork prints.
This work involved sensitive information. Employer systems and data are not shared here.
My focus is the whole workflow: the input, the repetitive steps, and what needs to happen next.
OCR, AI-assisted work and connected document actions, informed by my workplace experience.
Personal projects exploring research, scoring, drafting and application preparation.
Project work around capturing inquiries and preparing the next step for a business.
Public resources that make workflow ideas easier to inspect, adapt and discuss.
An interactive example of how I think about automation: a clear starting point, connected steps and a defined next action.
What starts the task, who does it, and where does the same work repeat?
What information moves between steps, and which decisions need a person?
What is the smallest useful workflow we can create within the agreed access and data boundaries?
How will we check the result, handle exceptions and decide whether it is ready to use?
Explore the connections. Drag the nodes to see how a task can move through preparation, review and action.
Explore my own projects, a receptionist demonstration and public resources for practical automation.
A lead-handling project for Synthetic Minds that brings inquiry capture, reply preparation and follow-up organization into one workflow.
View intake formA project combining job research, scoring and proposal preparation. The related public rubric shows how I approach deciding which opportunities deserve attention.
Read scoring rubricA command-line tool for reading a résumé, finding and scoring jobs, and preparing tailored application material.
A receptionist demonstration using Retell and Cal to explore appointment availability. Availability checks have been tested.
An unused, unpublished prototype connecting video ingestion with a research layer for newsletters, daily reels and YouTube videos, plus a connected Hermes agent for content repurposing.
I’m interested in the work between the tools: how information arrives, what happens to it, and what the next person needs.
These are the questions I bring to a workflow conversation. The right implementation depends on the work, the available access and the information involved.
What starts the task, who does it, and where does the same work repeat?
What information moves between steps, and which decisions need a person?
What is the smallest useful workflow we can create within the agreed access and data boundaries?
How will we check the result, handle exceptions and decide whether it is ready to use?
For AI consulting or implementation, contact me through Synthetic Minds. For professional background and opportunities, connect with me on LinkedIn.