$ whoami
Amit “Khivi” Khivesara · Fractional CTO
A 30-year engineering leader who’s actually hands-on with AI.
Sets strategy, writes the code, modernizes the stack. Fractional CTO for founders who need both.
$ cat strategy.md
Customers invest in outcomes, not technology.
I keep complexity in the business domain, not the stack — simple code, clear interfaces, boring technology a team can actually master. I design for change, not perfection: decisions that keep options open, the irreversible ones delayed to the last responsible moment, the cost of iteration kept low. When the assumptions change — and they will — the system bends instead of breaking.
The tradeoff is real: I'll hold a decision open longer than most are comfortable with, living in the ambiguity until it has to resolve. That patience is the thinking behind the Chief Data Officer Award I earned at Target.
$ git push origin main
Strategy without execution is a slide deck.
I build execution cultures: sprints that actually work, PRs that get real reviews, master always deployable, releases that are zero-touch. I codified it into The KHIVI Test — ten questions that tell you whether an engineering org is shipping or just busy. I've run this as CTO at two startups, where there's no slack for motion that isn't progress.
Small experiments over grand plans, fast feedback over perfect specs, teams that own their quality end to end. Product, engineering, and business aren't separate functions — they're one team solving one problem.
$ git log --author=mentor
The best engineers were mentored, not born.
The best engineers I've worked with weren't born great — they were mentored well. I push engineers to find their own solutions, walk them through the tradeoffs, and build environments where they can take real ownership.
I wrote Culture Seeds to capture the patterns that build those teams: how to hire, how to set standards, how to make quality the path of least resistance.
$ claude
AI made me faster, not looser.
Call it vibe coding if you like — I treat AI like any other part of the stack: deliberately. AI writes the code; I own the system. Tests stay non-negotiable, reviews stay real, nothing ships on vibes alone.
So I built a platform around it, not a prompting habit: dozens of skills that encode my workflows, specialized review and security agents, multi-agent orchestration through Cockpit, the open-source orchestrator I wrote, file-based memory that persists what matters, and a token-proxy CLI that cut my AI spend by more than half. Same instinct as the rest of the work — optimize the system, not the parts.
$ git log --since=1996 --stat
Three decades of doing it.
Bell Labs to Target to a roster of founders — the same range, applied across thirty years. The advisory work below isn't a footnote; it's the proof I already do the fractional-CTO job. Still shipping.
selected outcomes
10× the read performance of any commercial database — DataBlitz, taken research → worldwide deployment in Lucent products.
Deep learning shipped into one of the world's largest e-commerce search platforms — target.com relevance at scale.
A quant trading firm's stack stood up as employee number one — strategies and infrastructure trading US and Canadian equities, profitable daily.
$ git log --since=1996 --oneline
fractional & advisory: Needl · Dekco · Rezolve AI · Verne Global · MPower Plus — concurrent engagements, not a footnote.
$ history --since=1987
Testimonials · peers across the arc
"Amazing mentor. Khivi has helped many of us to shape our careers, towards becoming good developers. He has the notion of challenging engineers, allow them come up with solutions, guide them appropriately"
"Khivi looks beyond the code and beyond the technology to what is most important to a successful startup. Culture. Khivi helps to ensure that the companies he works with have the opportunity to maximize their potential"
"Khivi has such a mature, well-rounded perspective on product development and engineering practice. His point-of-view is paradoxically pragmatic yet strategic – allowing teams, roadmaps and architecture to stay flexible and nimble and not over-engineered"
$ cal book --with khivi
Let's build something that lasts.
Thirty years in, still in the code. Fractional CTO — technical enough to debug production at 2am, strategic enough to know what not to build.