vishesh / campaign reel
vishesh@work:~$

I build things people use, and make things people watch.

Final-year B.Com working across marketing, content, and go-to-market, who also builds with AI well enough to ship real systems. The film above is a fictional product campaign I wrote, shot, acted in, and edited myself.

$ open ~/neuromind --campaign
A campaign for a product that doesn't exist
Concept, brand, and creative, invented and executed solo.

NeuroMind is a fictional neural-implant brand I built a satirical campaign around. I created the brand identity, wrote the positioning, and produced the ad creative: scripting, storyboarding, acting in, shooting, and editing every piece. The campaign has a real point of view. It sells a brain chip while quietly arguing you don't need one. I also made an AI-deepfaked product-review video in the style of a well-known tech reviewer, doing the voice and face work myself.

The film at the top of the page is the centerpiece. Below is the brand world: the Instagram identity and campaign art, all designed for this project.

scriptwritingstoryboardingon-cameravideo-editingbrand-identityai-deepfakeart-direction
NeuroMind Instagram brand grid
NeuroMind brand world. Instagram identity, campaign art, and product visuals, designed for the project
$ ./run optimus --sre-agent
An autonomous AI system, built in 8 days with zero code background
AI Engineering Internship · Optimus Information Inc.

I joined Optimus with a marketing and content background. I had never written code, never opened VS Code, never touched an API. The brief was one line: build something useful with AI. Eight working days later I had shipped SRE Agent, an autonomous incident-response system that watches JIRA for critical tickets, pulls logs from Azure, runs a structured Claude diagnosis, and posts a full root-cause analysis back to the ticket in under 60 seconds.

I didn't write the Python by hand. I directed AI tools to build it, which is the actual point. The skill here is turning a vague problem into a working system by thinking clearly and directing AI precisely, then having the judgment to catch it when the output was wrong.

eval score across 7 scenarios0.89 / 1.00
full AI diagnosis per incident< 60 sec
total build time8 days
live production integrations3
python (ai-built)claude-apijira-cloudazure-app-insightspostgresql/neonprompt-engineering
SRE Agent system architecture
System architecture. JIRA detection, log aggregation, Claude diagnosis, decision gate, database

I documented the whole thing: a write-up on how a non-engineer ships a system like this with AI, the final presentation, and the full system design.

$ ./run ~/tools --live
Tools I built, live on this site
Most marketers buy their stack. I build mine.

AI is changing how customers find brands and how copy gets judged, so instead of writing about that shift I built tools for it and shipped them on my own infrastructure. Two kinds live here, all free, no signup. This is the part of "marketer who builds" you can actually click.

$ cat ~/oimedia --gtm
Go-to-market & product work for a B2B SaaS
GTM / Strategy Internship · OiMedia (SignifAI & ComplAI)

Over ~135 hours at OiMedia, a retail-tech SaaS company, I did the go-to-market work the marketing function actually runs on. I built an Ideal Customer Profile framework combining firmographic, operational, and psychographic segmentation; produced TAM market sizing; contributed to a Salesforce AppExchange listing; ran field research with independent retailers across Bengaluru; and authored a set of white papers on retail compliance and OOH media.

I also designed ComplAI, a mobile-first field-audit product, in Figma, mapping the full photo-capture, checklist, and auto-report workflow for on-site auditors.

icp-frameworktam-sizingmarket-researchpositioningfigma/uxwhite-paperssalesforce-appexchange
ComplAI mobile compliance flows in Figma
ComplAI, mobile-first audit flows in Figma
White papers authored for OiMedia
White papers authored on retail compliance & OOH media
$ stat ~/growth --audience
Content & audience growth
Built a real audience from zero, and got brands to pay for it.

I grew a gaming YouTube channel to ~10,000 subscribers and 851,000+ views as a teenager, learning audience retention, SEO, packaging, and consistency by doing it. The channel earned real revenue and attracted paying sponsors. I negotiated and integrated the brand deals myself, and my sponsored segments held retention at or above the channel average.

subscribers, grown from zero~10K
lifetime views851K+
watch-time hours43.1K
brand sponsorshipsdelivered, paid
YouTube channel analytics
Channel analytics: views, watch time, revenue
Audience retention graph
Audience retention across a sponsored video
Single video analytics in YouTube Studio
Per-video performance in YouTube Studio: 15.3K views and 1.1K watch-hours on a single upload
$ ls ~/work --shipped
Selected work
Design and web work shipped for real organizations.
National Digital Freight Exchange live site
web · webflow · for SemiCab × CII

National Digital Freight Exchange

Designed and built this live multi-page B2B site for a CII supply-chain initiative. Hero, data-led value proposition, whitepaper, and navigation, all in Webflow.

view live site
Youth Brigade donor pamphlet
collateral · print & social

Youth Brigade donor collateral

Redesigned the donor-facing pamphlet used to raise funds, plus a set of social posters. Information design built to persuade potential donors and investors.

Teaching financial literacy to students, ReapBenefit
Teaching financial literacy to students · ReapBenefit, Bengaluru
$ which tools
What I work with
marketing & gtm
  • content strategy
  • go-to-market
  • icp & segmentation
  • market research
  • positioning
content & creative
  • scriptwriting
  • video editing
  • on-camera
  • storyboarding
  • audience growth
ai & building
  • prompt engineering
  • ai-assisted dev
  • claude / claude code
  • automation
  • reading architecture
design & tools
  • figma
  • photoshop
  • webflow
  • premiere pro
  • notion