Two bartenders walk into a startup… and start rebuilding how people learn, earn, and invest in themselves. Dakota and Vladimir Kim met behind the bar in California, bonded over AI and entrepreneurship, and are now mixing hospitality, machine learning, and a new idea: treating people like startups and rewarding growth with “XP” that converts to real stakes.
TL;DR
Dakota (WA → CA) went from a childhood Fannie Mae flop to crypto trader → ML‑driven automation (Hydra).
Vladimir (Uzbekistan → NYC → CA) pivoted post‑COVID from hospitality to full‑stack software and ML, while staying grounded in people and process.
They’re co‑building a bartender app that starts as a visual “bar book” and expands into training, jobs, social discovery, and gamified career leveling.
Big thesis: Invest in humans like startups—contributors earn XP for real progress; backers earn recognition, returns, and reputation for funding human potential.
Deep Dive Summary
The Conversation
1) Origin stories: grit, reset, and the hospitality edge
Dakota grew up in Washington State in a family of general contractors (“brick and mortar first”). Financial curiosity hit early—he made his first investment at age 10 (Fannie Mae) and promptly lost the $300 he’d put in through his mom. That sting planted a seed: long‑horizon, fundamentals‑first thinking (Roth IRAs, compounding, “automatic millionaire” habits), later interrupted by the wild west of early crypto exchanges. By 2016–2017 he was trading crypto, studying technical analysis, and seeing 100% price spikes on thin‑liquidity exchanges with little or no KYC (“Know Your Customer”). He rode 2021’s run‑up, took profits, stepped away for two years, and is now back—this time aiming to remove as much human bias as possible via automation.
Vladimir was born in Uzbekistan in 1991, moved across the former Soviet region, and immigrated to the U.S. at 17. Post‑COVID closures in New York pushed a rethink: relying on tips from 20‑somethings at a club wasn’t a stable career. He retrained, first self‑taught (HTML/CSS/JS via FreeCodeCamp), then a full‑stack bootcamp, and now community college math (calculus, linear algebra) to prep for machine learning coursework and a UC transfer. One lesson he carries: being labeled “smart” can backfire; discipline beats talent, and preparation beats vibes.
Both men work in hospitality in California (they met at Blue Water Grill). That setting sharpened their core advantage: people skill at scale. When you serve 100 guests a day, you learn to read, adapt, and retain—skills most technical teams overlook until it’s too late.
2) What they’re building
Hydra (Dakota)
Hydra is a scanner/automation stack for crypto markets: multi‑source data collection (exchange APIs, order flow, volume, order book), Redis caching, ML‑based source selection and signal scoring, confluence analysis, multi‑timeframe filtering—the works. The aim: move from defensive, fatigue‑prone discretionary trading to 24/7 systematic decisions where models do the heavy lifting and humans tune objectives and risk.
(A note from the chat: they also touched on crypto’s messier bits—from early exchange “dust” and rounding to the general black‑box vibe of blockchain to retail—another reason to lean on rigorous data engineering and auditable logic.)
The bartender app (Dakota + Vladimir)
This started as a visual, ad‑free bar book (fast lookup during a rush), then expanded into a career and community platform:
Education & leveling: Games and drills that turn novices into competent bartenders (Michelin‑ready standards), with XP for mastering techniques, speed, and hospitality.
Jobs & reputation: Profiles, verified skills, and reviews; a way for managers to see true capability, not just a résumé.
Guest experience: A consumer mode that builds cocktails by flavor (sweet/sour/bitter/salty/spicy), cuisine pairing, or vibe—then hands the bartender a clean spec.
Social discovery: Find the bartender you like (not just the bar). See who can make your drink well. Keep people in the loop so the product doesn’t need to shout—community is the distribution.
They’re also exploring health‑aware mixology: integrating herbs/spices (turmeric, ginger, rosemary), ferments, and other functional ingredients; and using AI to personalize choices. One provocative idea from the chat: ethically collecting face images + favorite cocktails to explore whether models can map phenotypic cues to preferred flavor profiles or lighter‑impact options (strict opt‑in, privacy‑first). Whether or not that specific line bears fruit, the broader point stands: use data to make indulgence a bit smarter.
3)The Trillion Dollar Idea - Investing in People like Startups: The XP Economy
Their most interesting idea reframes education and early contribution:
Contributors (coders, marketers, researchers) earn XP for real work.
XP converts into an equity‑like stake or revenue share if/when a product hits.
Investors can back people, not just projects, and earn reputation and returns for funding human potential.
To keep momentum, non‑contributors dilute slowly—a nudge to stay engaged or re‑up.
The same game layer works for backers: badges, totals, and “win rate” signals let investors display the breadthand quality of their human investments.
It’s not about chasing quick exits; it’s about structuring incentives around growth. The theme repeats across their work: gamify the right behaviors, track real progress, reward contribution, and make the system feel fun enough to keep people in it.
4) Grit, activation energy, and lowering internal resistance
The discussion kept landing on motivation mechanics:
“Big T” trauma often fuels entrepreneurs—chips on shoulders that become engines.
Activation energy matters: most people fail at the start line. Lower the slope (make the first reps fun), and you change outcomes.
Pain tolerance is trainable. In biology terms, lowering internal resistance (inflammation, stress) via sleep, breath, cold, movement, light, and nutrition increases the capacity to keep going.
The karate‑kid principle: disguise hard work as something else (“wax on, wax off”) so beginners accrue wins before they notice the burn.
They even bring this home in parenting: consistent routines, graded reductions in “sleep crutches,” and modeling behaviors (do push‑ups in front of your kids; don’t just talk about them). As Vladimir put it, “Your kids watch you grow up too.”
5) Content is king (and why they’re recording)
A recurring motif: content drives intelligence. The early LLM breakthroughs exploited sentiment signals in large text corpora; the same logic applies at smaller scales. Recording conversations, capturing decisions, and storing patterns lets you train agents that reflect your real judgment—Steve’s “AI dad” idea, but for founders, bartenders, and investors. If you want machines to help, feed them the right stories.
Practical takeaways
Blend tech with touch. The best builders can talk to strangers, retain customers, and still ship.
Gamify contribution, not just consumption. XP for real progress keeps collaborators and backers engaged.
Design the on‑ramp. Reduce friction for the first five minutes; the rest takes care of itself.
Invest in humans. Reputation systems for backers and contributors could rebalance early‑stage opportunity.
Record your process. Today’s notes are tomorrow’s model weights.
In Closing
If the future of work is fewer résumés and more receipts, Dakota and Vladimir are building the receipt printer. One project makes markets legible to machines; the other makes careers legible to people. Both rely on the same principle: give the right effort a score, give the right people a stage, and let value compound—behind the bar and far beyond it.
For a music clip, this is from a recent session ‘For What It’s Worth.’ A great example of a music collaboration between several different people with different perspectives and musical tastes coming together and all enjoying it! Grant (Vocals/Guitar), Andrew (Guitar), Randi (Vocals), Tammy (Vocals), Alan (Bass), and Johnny (Conga) As usual, there is a brief Substack mute for some reason at the beginning.
Share this post