Valentyn Yakovliev — Full-Stack Product Engineer

In 10 seconds

The short version

Experience

8+yrs

Shipping web, mobile and desktop products — from seed-stage startups to enterprise B2B SaaS, with 2+ yrs of AI, LLMs and agent systems in production.

What I do

0 → 1

I take ideas from blank page to live, paying product — design, code, launch, and iterate, with a fleet of AI agents doing real work alongside me.

Right now

AI-first

Shipping AI features and agentic workflows in production, and running my own SaaS with agents for SEO, content, QA and social media.

What I can do for you

Plain English

Build a product from scratch and ship it.

Give me a Figma or just an idea. I write the backend (Node.js or Python/FastAPI), build the screens, wire up payments, deploy it, and keep it running. No handoffs, no gaps.

ProofBuilt a full-stack SaaS pet project solo — launched, real paying users, web + native Android.

Be the only engineer a small team needs on the frontend.

One person who decides the architecture, writes the code, reviews the pull requests, talks to designers and backend, and unblocks everyone. Senior-level judgment without the politics of a big team.

What it looks likeAt Oboard I was this person for 4 years — reviewed every frontend change, owned the design system, and shipped the product into Monday, Salesforce, Jira and Confluence from one codebase.

Own a whole feature area, end to end.

From first conversation with stakeholders, through design and code, to QA and production. I talk directly to founders and C-levels — not just to engineers.

ProofOwned the OKR dashboard at Oboard for 4 years; reviewed 100% of frontend changes.

Add AI to your product without it being a gimmick.

I've shipped real AI features and agentic systems (OpenAI, Claude, Gemini, LangChain, LangGraph, Agents SDK) — from search and chat to agents that test features, write docs, create content and run social media. Plus the infra around it: queues, caching, cost control, observability.

ProofRAG system on Qdrant + multi-provider LLM routing, and a fleet of internal AI agents doing production work.

Make one app work everywhere.

Web, iPhone, Android and desktop from one shared codebase — cheaper and faster to maintain. I use Capacitor, Ionic and Electron so the same project ships to browsers, app stores and desktops without rebuilding from scratch.

ProofShipped native iOS & Android apps from one codebase (Capacitor), both live on App Store and Google Play, with desktop builds via Electron.

How I ship fast without cutting corners

My method · SDD

Spec-Driven Development

I built my own SDD framework on top of GitHub Spec Kit — and I ship with it.

Most teams either over-plan and never ship, or ship fast and break things. My framework forces every feature through a short, structured spec first — then AI coding agents (Claude Code, Codex, Windsurf, Devin) generate the scaffolding, tests, and first pass of code against that spec. Result: faster delivery, fewer bugs, real docs from day one.

This isn't theory — I dogfooded it on a personal SaaS project and accelerated feature work at my day job. Same framework, two different contexts.

github.com/VaiYav/speckit-product-forge →

01SpecWrite the contract — what, why, edge cases↓

02PlanAI breaks spec into tasks + file layout↓

03GenerateScaffolding, types, tests — all from spec↓

04ImplementI write the real code against generated tests↓

05ShipSpec is now the docs — nothing goes stale✓

Built with it

Personal SaaS

Pet project, shipped end-to-end — web + native Android.

Applied at

Soulwise AI

Features shipped faster with tests written first, not last.

Open source

speckit-product-forge

Extending GitHub's Spec Kit for real product teams.

Skills, honestly

What I'm actually good at

CoreExpert · Daily use

I've built real products with these. Ask me anything.

Vue.js 2 / 3 8 yrs Nuxt 3 / 4 4 yrs TypeScript 5 yrs JavaScript / HTML / CSS 8 yrs Composition API Pinia / Vuex Responsive / Mobile web

StrongShip-ready

I've used these in production and would again tomorrow.

NestJS Node.js Python / FastAPI React Next.js MongoDB PostgreSQL Redis REST & WebSockets RabbitMQ / Kafka BullMQ / Queues Docker & CI/CD Stripe / Payments Capacitor (iOS / Android) Ionic Electron (desktop) OpenAI · Claude · Gemini LangChain · LangGraph Agents SDK RAG · Qdrant Prompt engineering Testing strategy · Vitest · Playwright · E2E TDD / BDD SDD / Spec-Driven AI coding agents (Claude Code, Codex, Windsurf, Devin) Product & UX design Sentry / New Relic / PostHog / Amplitude FSD · DDD · EDA · Hexagonal · Event-driven Technical leadership Mentoring & code review 0 → 1 launch App Store & Google Play

AI & GrowthAgents · Marketing · Scale

I design AI systems and growth loops that move the needle.

AI agent design & creation Multi-agent orchestration Human-in-the-loop SEO / AEO / GEO ASO i18n / l10n · Localization at scale Content & copywriting Marketing automation Social media automation CRO / A/B testing Analytics & attribution

WorkingComfortable

I can pick up tickets here without slowing the team down.

Angular Svelte Quasar Monday.com / Salesforce / Jira / Confluence apps Micro-frontends

→ AnyFramework-agnostic

Vue is my daily driver, but the patterns transfer everywhere.

Strong Vue.js and React background transfers directly to Angular, Svelte, Next.js or Nuxt — same component model, same state patterns, same testing approach. I've switched before, I'll switch again.

Career at a glance

2017 — Today

Past Present

'17

'18

'19

'20

'21

'22

'23

'24

'25

AI Consultant · Freelance

Consumer AI Platform · Full-Stack

Oboard · Senior Frontend / Feature Owner

AnyforSoft

IT Svit

Codica

QPARD

Current (incl. parallel consulting) Long-term (4 yrs) Mid (1–2 yrs) Short-term / contract

Experience in detail

Most recent first

2025 — Present

Ongoing · freelance

AI Integration Consultant

Independent · 3 clients · Helping product teams adopt AI without the hype

Delivered end-to-end AI adoption roadmaps for 3 companies — from LLM provider selection and cost modelling to phased rollout plans tied to real engineering capacity.
Built dependency and responsibility matrices for each client's AI stack — clearly mapping ownership across model provider, infrastructure, data pipeline, and product layer so nothing falls through the gaps.
Designed and prototyped internal AI agents for dev workflows: automated code review assistants, documentation generators, sprint planning helpers — practical tools the teams actually use.
Consulted on RAG architecture, prompt engineering, and multi-provider fallback strategies (OpenAI, Claude, Gemini) with a focus on cost control and reliability in production.
Delivered practical integration playbooks — structured guides that let teams ship their first real AI feature in 2–4 weeks without disrupting existing workflows.

LLM Integration AI Agents LangChain · LangGraph RAG · Qdrant OpenAI / Claude / Gemini Tech Roadmapping Consulting

2025 — Present

1 year

Full-Stack Engineer

Soulwise AI · AI-powered consumer platform · 10k+ users

Made the app 36% faster to load (bundle shrunk from 250 KB to 160 KB) by splitting code smarter and loading translations only when needed.
Cleaned up the codebase: full TypeScript strict mode across the project; reorganised the code into 14 clear feature modules and deleted 3,500+ lines of duplicates.
Shipped native iOS and Android apps using Capacitor with one shared codebase — both live on App Store and Google Play.
Built real AI features connecting OpenAI, Anthropic Claude, and Google Gemini, with a vector-search layer (Qdrant) so the AI has real context, not just vibes.
Real-time updates and notifications via WebSockets and background queues (Redis + BullMQ), so users see changes instantly.
Set up the testing + CI quality gates — Vitest for unit tests, Playwright for full end-to-end — so broken code can't reach production.

Vue 3 Nuxt 3/4 NestJS Python FastAPI TypeScript MongoDB Redis Capacitor OpenAI / Claude / Gemini LangChain · LangGraph Qdrant Vitest Playwright

2022 — 2025

~4 years

Senior Frontend Developer · Feature Owner

Oboard · B2B SaaS for company goal-setting (OKRs) · Team of 2 frontend + 5 backend engineers

Owned a whole feature from idea to launch: the OKR dashboard and analytics. Talked directly with C-level execs about what to build next.
Modernised the frontend across several major framework upgrades; was the primary reviewer on every frontend change for quality control.
Shipped the product inside other platforms — Monday.com, Salesforce, Jira, and Confluence — so customers could use it where they already worked.
Built the build pipeline that produces 5+ versions of the app from one codebase: standalone webapp, Monday embed, Salesforce app, Atlassian apps, and mobile.
Integrated payments and subscription billing.
Built 3+ internal tools for the QA and dev team during sprint gaps — cut repetitive manual work by roughly 40%.
Mentored and reviewed across the frontend team; partnered with 5 backend engineers on API shape and data contracts.

Vue 3 Nuxt 3/4 TypeScript Quasar Pinia

Feb — Apr 2022

3 months · contract

Frontend Developer (Vue.js)

ITOMYCH Studio

Vue.js

Jul 2021 — Feb 2022

8 months

Frontend Developer (React, Vue)

AnyforSoft

React Vue.js

Jun 2020 — Jun 2021

1 year 1 month

Frontend Developer (Vue.js)

IT Svit · Full-time

Vue.js

Mar — Jun 2020

4 months

Frontend Developer

Freelance · Remote

Aug 2018 — Mar 2020

1 year 8 months

Frontend Vue.js Developer

Codica

Vue.js

Aug 2017 — Jul 2018

1 year

Frontend Vue.js Developer

QPARD

Vue.js