Strategy and focus
I help you pick the one thing that matters. The rest can wait. Or never happen at all.
personal site · 2026 edition
DmitriiSergeev
Building AI-first companies.
founder · creator · advisor
Building SaaS since 2015. Took Carrot Quest from zero to $3M ARR. No investors, still 35% growth a year.
Building Dashly. The multichannel AI SDR for inbound. Three AI agents and two scenarios work the same lead across web chat, WhatsApp, Telegram, and email. Engage, qualify, and book every inbound lead 24/7. Double your pipeline without hiring a single SDR.
/ 01 / bio
Marketing, product and sales. One system.
In most companies product, marketing and sales live on different floors. I never worked like that. At the center of all three sits a conversation with the customer. That's why I build them together, and why I push teams to build bridges between each other. In my first company I shipped the product, ran sales, and wrote the copy myself. Never unlearned it.
My first business was Anivers, a web studio I started in 2009. Clients were Sberbank, Rosneft, Imperial Tobacco. I left enterprise once it was clear it doesn't scale and burns the team. Carrot Quest came next. Now Dashly.
AI-first in the product, AI-first in the process. I write code, landing pages, decks with Claude. The artifacts that fall out of that stream live here.
Right now I'm Founder & CEO Dashly, Founder & Advisor Carrot Quest.
/ 02 / numbers
01$3MARR at Carrot Quest0 → 1000+ customers, no funding, 35% growth a year.
0210+years in SaaS & AIWrote code in 2009. Ran a team in 2015. Writing code again in 2026.
0350kreaders / monthB2B blog I built at Carrot Quest.
042.5kTG subscribersChannel on going from 0 to 1.
/ 03 / trajectory
Seventeen years. One line.
- 2024 → nowFounder & CEO · Dashly.ioAI agents for inbound · $252K → $1M ARR goal
- 2024 → nowFounder & Advisor · Carrot QuestCo-founder · scaled to $3M ARR
- 2023 → nowAngel / Mentor / Advisor · IndependentC-level in EdTech & MarTech
- 2024 → nowSolo maker · AI toolingGPT agents, voice bots, no-code
- 2015 → 2024CEO · Carrot Quest0 → 1 → product-market fit
- 2009 → 2015CEO · Anivers StudioEnterprise: Sberbank, Rosneft
/ 04 / artifacts
HTML, decks, essays.
Made with Claude.
An artifact is a thought caught before it drifted off. Five areas I work in: CEO, marketing, product, AI, vibecoding.
01 · VibecodingEssay
Three sets of sales skills for Claude: compared
What's inside, for whom, when to use, what to adapt.
A neutral comparison of three open-source sales-skill sets for Claude Code (sales-skills/sales, zubair-trabzada, louisblythe): what's inside each, who it's for, when to install it, and what to adapt. A head-to-head table plus a “which set for which situation” matrix.
2026 · v 1.0read →
02 · AITool
Grading AI skills in hiring
G0→G3, 5 dimensions, 17 questions. No artifact, no offer.
How we grade candidates' AI skills: four grades G0–G3, five dimensions (Delegation, Description, Discernment, Building, Diligence), per-role bars, 17 questions, and a live practical block. Built on the Zapier v2 rubric and Anthropic's 4D framework. The rule: no artifact (prompt, automation, agent) and no number — no offer.
2026 · v 1.0read →
03 · CEOTool
Hiring A-players: process, evaluation, onboarding
12-stage funnel. ABC matrix. AI inside, the call stays human.
A full hiring workflow: kickoff, sourcing, ABC-matrix evaluation (the “Who” method by Smart & Street), interviews, offer, and onboarding. 12 stages — now with AI inside: candidate scoring, auto-linking records, prepping onboarding in minutes. The final call always stays human.
2026 · v 1.0read →
04 · AITool
Meeting pipeline: Fireflies → Telegram
Every meeting routes itself to the right team
Record a call in Fireflies, then zero manual work. A cron every 15 min classifies transcripts, files them into per-team repos, and pings Telegram. Unsure → it asks with one tap; a Cloudflare Worker on a Claude tool-use loop instantly routes the meeting or saves a rule.
2026 · v 1.0read →
05 · CEODeck
I have an idea. B2B SaaS. What now?
Traction map: from idea to $1M ARR. StartupCamp Dilijan.
Go-to-market for B2B SaaS, drawn from building Carrot Quest and Dashly. Seven stages from idea to $1M ARR — each gated by one concrete artifact. The 4 mistakes engineering teams make, what an MVP really is, and exactly what I'd do by hand in 5 days from scratch today.
2026 · StartupCampread →
06 · VibecodingTool
Zero to commit. A hackathon participant guide.
Mac and Windows, mixed Git experience.
Set up GitHub, Cursor, and Claude Code in 30 minutes. Team workflow in one repo: branches, PRs, conflicts. Deploy to Vercel. Includes a downloadable Claude system prompt that walks a person through the setup step by step.
2026 · v 1.0read →
07 · MarketingEssay
Content Hubs. A Field Guide.
Four models. Eight rules. Anti-patterns.
What a content hub is, and how it isn't a blog. Four models: Hub-and-Spoke, Content Library, Topic Gateway, Content Database. Eight rules that hold a hub together. Eight anti-patterns that break it. A decision tree for picking the right model.
2026 · v 1.0read →
08 · ProductTool
Dmitrii Ive. Design System v 1.0.
Colors, type, components, voice.
Full spec: 10 color tokens across two themes, three font families, six components, and a copyable starter CSS. Take it, and build a new project in the same visual language.
2026 · v 1.0read →
09 · VibecodingEssay
Claude Skills. A conceptual guide.
What it is, when to use one, how it's wired.
Skills sit between prompts, project knowledge, and MCP. A folder of instructions Claude loads only when the task matches. Six situations where a skill pays off. What to do and what to skip.
2026 · v 1.0read →
10 · VibecodingEssay
Vibecoding with Claude Code
Six observations from a CEO
How I build and rebuild whole products in a conversation with Claude. Context beats prompts, sessions beat one-off queries, and where it all breaks.
2026 · v 0.1read →
11 · AIDeck
AI SDR: funnel economics
12 months of lessons from Dashly
$252K ARR, $1M target by 2026. What works in inbound funnels powered by AI agents. And what doesn't.
Coming soonread →
12 · CEODeck
0 → $3M ARR with no funding
Carrot Quest, a ten-year retro
Bootstrapped, profitable, 35% YoY. What I'd repeat, what I'd skip, and why a CEO keeps working after the exit.
Coming soonread →
13 · MarketingDeck
Inbound funnel teardown
From SEO to handshake
How a blog with 50k readers/month becomes SQLs, then revenue. Content, scoring, weekly rituals.
Coming soonread →
14 · ProductDeck
Product Research · JTBD as craft
Based on Product Research Book
A compact take on my product research book. Custdev, hypotheses, finding the real pain.
Coming soonopen ↗
The catalogue keeps growing. New HTML artifacts go into public/artifacts/ with an entry added to lib/artifacts.ts.
/ 05 / how I help
Partner. Mentor. Advisor.
Teams that drive themselves
What changes when the CEO isn't in every meeting. Roles, rituals, culture. None of the HR fluff.
Marketing that pulls leads
Inbound that actually works in B2B SaaS. Positioning, content, traffic. The same stack pulled 50k readers a month at Carrot Quest. I'll show you how it's wired.
Sales that close
Your first sales hire and their KPIs. Why marketing and sales still don't talk in your company. What syncs them with two recurring meetings a week.
Product-market fit
Custdev without self-deception. Hypotheses you can run in a week. Channels where customers are already paying.
/ 06 / writing
Channel on going from 0 to 1.
Telegram · ive_from_0_to_1 ↗
My book on product research. JTBD, custdev, hypotheses.
Product Research Book ↗
spoke on stages
Netology · EcomExpo · FRII · RIF · Product Mindset · Product Camp · ScrumTrek · Epic Growth
/ 07 / connect