Development Project - Automations for deep tech entrepreneurs

Overview

Inspired by the NFX article, “The 3-Person Unicorn Startup,” this project's goal was to develop a modern skillset for scaling and optimizing my own workflows. With input from current DSV founders, I defined 15 common workflows that a deep tech entrepreneur might conduct regularly. I then performed a landscape analysis of available tools that could be integrated to automate these processes, filtering for functionality and cost. The final output was a suite of 20+ custom automations and a new widely-applicable technical skill.

Featured automations

‣ Email qualified LinkedIn users to automatically schedule meet-and-greets

‣ Predict how a LinkedIn user may feel about a topic and suggest conversation starters

‣ Create LinkedIn post with citations about a topic based on example posts

‣ Find all employees of a company matching a job description and send connection requests

‣ Create SEO-friendly blog posts with citations and share to social media

‣ Share and repost stories from selected RSS feeds with comments/summary on LinkedIn

‣ Monitor for newly filed patents matching specific keywords or by certain orgs/individuals

‣ Monitor for research publications in a given subject and summarize findings

‣ Create first draft pitch decks using NLP and Notion databases

‣ Monitor for relevant grant opportunities and add due dates to Google calendar

‣ Create patent draft and perform preliminary Freedom-to-Operate analysis

Additional outcomes

Operationalizing within DSV: In an effort to apply my new automation skillset towards real problems encountered by founders, I worked with Tanya Mangoma, a DSV founder in residence, to help automate her customer discovery and outreach processes. Using Make.com and PhantomBuster, we set up multiple workflows allowing her to filter and target LinkedIn users by company and role and then send connection requests or individualized messages.

Custom GPTs: With OpenAI's release of custom GPT's, I can now call my Make automations directly from a GPT using the new “actions” feature, coupled with a Make webhook. This allows me to create Retrieval Augmented Generation (RAG)-assisted LLM tools with little to no code.

Portfolio site: To showcase this project and all my future work in the VSD program, I built myself a fully new portfolio site (the one you're looking at now) using NextJS, Tailwind CSS, Typescript, and Spline 3D. I integrated a Make automation into the site to manage the subscribe/unsubscribe functionality, and I plan to continually add unique features using my newfound automation and API integration abilities.

Key takeaways

Prior to this project, I had limited experience in workflow automation and API chaining. The skills that I gained through this work will enable me to continually optimize my workflows throughout my entrepreneurial career and keep me at the forefront of LLM-based tool integration.

Tools used