RESEARCH × AUTOMATION
A research loop that keeps going.
Exploring GitLab issues through an ongoing, automated research workflow—turning a sprawling backlog into something I can investigate.
✳ A BUILDER WITH A LOT OF QUESTIONS
I don’t just
try AI.
I tinker._
I’m Deepak. I turn curiosity about AI into working experiments, agent workflows, and lessons worth sharing.
Building a case for the AI Tinkerer role
DEEPAK’S WORKBENCH01 — 03
GitLab issue researchWORKFLOW SKETCH
INPUTFind the question
01EXPERIMENTGather context
HUMAN IN THE LOOPReview the evidence
03+ A repeatable starting point for deeper investigation.
Curiosity → experiment → understanding▮
Always something on the bench.
A FEW TOOLS
ON MY BENCH
ON MY BENCH
ChatGPTClaudeGeminiMuseOpenClawHermes
+ more01 / SELECTED EXPERIMENTS
Curiosity, put to work.
Less tool collecting.
More “what can I build with this?”
MULTI-AGENT × PLANNING
More perspectives. One useful plan.
A LinkedIn networking system where multiple agents exchange ideas and prepare a plan for me to review.
LOCAL × CLOUD
The same curiosity. Different machines.
Hands-on experiments with OpenClaw and Hermes across cloud environments and a Mac mini, alongside everyday AI tools.
Workflow illustrations shown above. Screenshots, demo recordings, and project links are coming next.
02 / FROM EXPERIMENT TO EXPLANATION
The part after
“that’s pretty cool.”
A new model is a starting point.
A useful lesson is the goal.
01
Find a real question.
Start with a task someone actually wants to do. Give the experiment a purpose before picking a tool.
02
Get my hands on it.
Try the workflow. Change the setup. Ask where it helps, where it breaks, and what still needs a person.
03
Make it teachable.
Turn the moving parts into a clear demonstration: what to try, what to expect, and what to watch for.
A FIRST LESSON I’D MAKE
What changes when agents work together?
One task, multiple perspectives, and a walkthrough of the plan they produce.
dg✳
DEEPAK REDDY GATHPACurious by default.
const mindset = { curiosity: "always on", nextStep: "try it"}
03 / THE PERSON BEHIND THE PROMPTS
I like figuring
things out.
I’m Deepak, a computer science graduate student at UCF, open-source contributor, and someone who keeps finding another AI experiment to try.
My work moves between code and curiosity: GitLab contributions, automated research, collaborating agents, and local setups. I care about understanding the machinery well enough to explain it.
That’s what draws me to the AI Tinkerer role: the chance to turn hands-on exploration into something people can learn from.
✳ LET’S BUILD SOMETHING WORTH EXPLAINING
A lot of curiosity.
Ready to put it to work.
I’d love to bring my experiments, engineering mindset, and appetite for learning to your AI education team.