DEV Community: Val Neekman

Experience is making a lot of mistakes in a narrow field.

Val Neekman — Sat, 14 Feb 2026 00:28:32 +0000

Experience is making a lot of mistakes in a narrow field.

The seniority we take pride in is simply the accumulation of those mistakes, and the lessons we chose to learn from them.

We improved because we could walk over to an experienced engineer’s cubicle, ask questions, and get unstuck.

Now it’s time to pay that forward.

In the AI era, though, the weight of “old” experience is slowing many senior engineers down. Some are fixed in their tracks, unwilling or unable to adapt.

If my experience isn’t transferable to the generations that come after me, then it isn’t worth much.

When companies abandon new graduates, they’re abandoning the next generation - our future.

Can I do my part?

Yes. Even if nothing else, keep your resistance to yourself. If you don’t like AI or feel incapable of adapting, don’t discourage the young ones.

Let them build. Let them explore. Let them surpass us.

I’m doing my part by building ehAye Engine (https://ehAye.io), a voice-first platform that mentors younger developers through natural conversation, reaching each of them virtually.

Code coverage is not a one-size-fits-all rule.

Val Neekman — Mon, 02 Feb 2026 18:54:35 +0000

The obsession with “85% code coverage” usually comes from people far removed from real software and systems architecture. It is a checkbox mindset, often driven by me-too thinking rather than engineering judgment.

If your code runs close to the silicon, has no presentation layer, and is mostly deterministic logic, then yes, 60–70% unit test coverage is solid. Push beyond that and diminishing returns kick in fast.

In large projects, especially outsourced ones, coverage metrics get gamed. A function gets called, a test passes, the percentage goes up. Management cheers, cookies are shared, and nothing meaningful is actually validated.

For any software with a real UI, unit testing beyond ~40% is usually a tax on the project’s bottom line.

What has worked extremely well for me:

Any logic that can be turned into a pure function has 100% unit tests.
Anything that can be mocked gets mocked.
Everything else moves to end-to-end tests.

E2E testing used to mean brittle browser automation scripts. That era is over. Today, I run full end-to-end flows using my own product, ehAye Engine, with built-in browser automation. Sessions are fully recorded, screenshots captured per page and tab, anomalies flagged.

First pass is read-only. Second pass allows mutation.

Before any deployment, staging (identical to production, but private) must pass full E2E to unlock upstream merge. No exceptions.

I am perfectly comfortable with 20% unit test coverage when I know E2E is validating 100% of the system behavior.

The key is measurement, but let ehAye Engine do the heavy lifting. Record everything, capture screenshots, narrate what is happening. Repeatable, fast, high signal.

Sit back, relax, and prosper.

Download at: https://ehAye.io
Watch video at: https://www.youtube.com/watch?v=Bb1C1NmduV8

ehAye Engine + Claude Code: Executives Finally See Product Quality, Firsthand

Val Neekman — Mon, 02 Feb 2026 18:50:34 +0000

Thanks to ehAye Engine and AI Agents -- for the first time, executives can directly observe the real quality of their product, end to end, exactly as customers experience it.

In this short walkthrough, we show how ehAye Engine combined with Claude Code enables live and recorded E2E testing with clear visibility, eliminating green dashboards, guesswork, and filtered reports.

Trust the metrics, but verify the reality.

Download at: https://ehAye.io

Watch video at: https://www.youtube.com/watch?v=C1hXYKOHHVI

7-minute coding session with voice with Claude Code

Val Neekman — Mon, 02 Feb 2026 18:48:45 +0000

In this 7-minute coding session, you’ll see how ehAye Engine transforms Claude Code from a text-box assistant into a natural, conversational teammate and mentor.

Instead of typing prompts and reading walls of text, you talk to Claude Code as it codes, speaks, explains, and guides you in real time.

The result is a more human way to build software, where you stay focused on the code and the system works alongside you, not inside a box.

Download at: https://ehAye.io

Watch the video: https://www.youtube.com/watch?v=_MQ-ycOt0Q4

Why ehAye chose Claude Code as its primary AI agent?

Val Neekman — Mon, 02 Feb 2026 18:45:17 +0000

There is a consistent pattern across organizations: people do their best work when they feel trusted, respected, and genuinely invested in the outcome.

When contributors are treated as long-term stakeholders rather than conditional participants, creativity compounds. An ownership mindset emerges naturally, driven by conviction rather than compliance.

At scale, these dynamics matter even more.

High-performing organizations rely on alignment and trust at the top. Strategic partnerships, capital support, and long-term backing often play a stabilizing role during early growth. Aligned external gravity can be the difference between survival and failure in formative stages.

History also shows another recurring lesson.

Stated missions only endure when incentives, governance, and execution reinforce them over time. When there is tension between origin, structure, and long-term behavior, credibility erodes and expectations reset.

Sustainable companies are built on positive internal energy, shared purpose, and confidence in leadership. Teams that trust the system they operate within stay engaged, resilient, and committed through uncertainty.

Product decisions reflect these values as well.

For example, some platforms choose to obscure internal reasoning for valid technical or commercial reasons. Others prioritize interpretability and transparency to support downstream builders. These choices directly influence how ecosystems form, how developers engage, and how much trust accumulates over time.

This perspective is why ehAye Engine chose Claude Code as its primary AI agent.

The objective is not just raw capability. It is clarity, transparency, and genuine bi-directional interaction, where users can understand how the system operates, collaborate with it naturally, and build with confidence.

If you want to interact with Claude Code through natural, two-way conversation, try ehAye Engine.

https://ehAye.io

Voice-First AI Coding Is Here

Val Neekman — Mon, 02 Feb 2026 18:42:26 +0000

What if your coding agents could speak to you while you work?

ehAye Engine adds voice, vision, and control to AI coding agents like Claude Code. Listen to responses, speak commands, and execute tasks - without breaking focus or switching between code and chat windows.

Humans evolved to listen while working with their hands. Constant visual context switching creates friction. ehAye Engine removes that friction by transforming silent AI into a focused, real-time coding companion.

You stay in your editor.
Your hands stay on the keyboard.
Your assistant speaks when it matters.

This is not another chatbot interface. It's an upgrade to how developers actually think and work.

Download ehAye Engine free at https://ehaye.io

🗑️ Garbage in - Garbage out 🗑️

Val Neekman — Tue, 27 Jan 2026 00:38:42 +0000

Using AI coding agents improperly is like finally getting the chance to ask out the girl of your dreams, then freezing because you don’t know what to say or how to say it. Disappointment is inevitable.

LLMs are not all equal. You get what you pay for. If you’re using a free version of something, don’t expect miracles.

Now assume you do pay for the top tier. You use the best available model, fire up Claude Code, and on paper you have everything required to build quality software.

Yet the results still fall short.

The reason is simple. You don’t know what to ask.

If you give Claude Code a half-baked specification, the output will be half-baked. No matter how strong the model is, it cannot read your mind. Text goes in, text comes out. Garbage in, garbage out.

The best tools combined with an imperfect specification is still a bad path forward.

What actually works is orchestration.

Start by creating a spec file. Ask Claude Code to spin up five agents, each an expert in a different area, to critique it, expand it, fill gaps, and surface edge cases. Have them add structure and Mermaid diagrams to show flow. Then have the main agent review their work and produce a finalized specification you can agree on.

Only then move to implementation. Bring back those same agents, assign each a section, and have the main agent enforce strict adherence to the agreed-upon spec.

Even at that point, most people hit another wall.

The computer beats you on stamina and perseverance. For every line of input you type, you get a page of output you’re expected to read, validate, and respond to. Before long, you’re lost between walls of text.

That’s exactly what happened to me while working on an AI-enabled SaaS project. Two weeks in, I was fatigued and discouraged. Not because the AI was weak, but because the interaction model was exhausting.

I caught myself thinking, I wish I could talk to my AI coding agents and have them talk back. Two weeks later, I had a prototype. It was addictive. My productivity jumped roughly tenfold. 🚀

That’s when I abandoned the original SaaS project and focused entirely on building ehAye Engine, which gives Claude Code voice, vision, and control.

Claude Code stopped being just a tool and became a teammate.

Once I was free of the text box and could have a natural two-way conversation with Claude Code, everything changed.

Mileage may vary, try it and report back. I’m curious how your view changes.

Val Neekman
CEO and Founder, Neekware Inc.

🚀 https://ehAye.io

I made Claude Code talk to me while it works

Val Neekman — Mon, 26 Jan 2026 17:29:31 +0000

I got tired of watching terminals scroll while my AI coded.

So I built ehAye Engine - a local-first system that gives AI agents voice, browser control, and system access.

What it does

Voice: Claude Code narrates what it's doing in real-time
Vision: Browser automation via MCP
Control: System access for file operations, E2E testing, automation

Why I built it

Coding: I'd ask Claude to refactor something, then just... watch logs scroll. Now it talks through the changes as it makes them. I can code eyes-free.

E2E testing: 25 pages, Stripe checkout, the works. Claude ran the whole suite while I made coffee - narrating every step. CEOs reactions: "Wait, I actually understand what it's doing."

That's when it clicked. AI shouldn't just be for devs reading terminals. It should be a teammate anyone can work with.

Demo

Coding with voice:

E2E testing hands-free:

Try it

Free to try: ehaye.io

Works with Claude Code, Cursor, Gemini, any MCP-enabled assistant.