Denis Mită

Software Engineer

I build polished web products from scratch — sharp eye for design, genuine curiosity for AI and developer tooling.

Experience

2024-2024 Stella
2023-2024 Try it on AI
2022-2023 Deepwaters
2021-2022 Minerva Gate
2017-2022 Blink
2016&2017 Google
2015-2017 CS Academy

About

Back in middle school, I discovered a passion for digital painting in Photoshop. In high school, my love for math pulled me into competitive programming. These two worlds — visual craft and algorithmic thinking — naturally converged into my career as a software engineer.
Since 2015, I've shipped products across startups and at Google, spanning distributed ledger protocols, web-wide identity and payments infrastructure, generative AI applications, and crypto asset trading platforms. More recently, I've developed a strong focus on AI tooling — I actively track the latest coding agents, LLM releases, and automation frameworks, and integrate them into my workflow to ship faster and smarter.
I care deeply about product quality — from pixel-perfect interfaces to clean architecture and seamless user experiences. My work is industry agnostic: what drives me is building things that feel right, with code that holds up over time.
Fun fact: "Denis Mită" translates to "Denis Bribe" in English — hence the handle "dbribe".
Scroll to see detailed work experience

Software Engineer

Stella
May 2024 - Jul 2024

Stella grew out of the team behind Try it on AI. The product was an AI-native digital marketing platform — combining social media management, collaborative workflows, and LLM-powered content tools customized to each client's brand voice and strategy.

As a full-stack engineer, I owned the setup of the platform's top-level pages and core sections. I integrated billing via Stripe and built several custom widgets from scratch, including a calendar scheduling tool.

The most impactful feature I shipped was a unified social media publishing interface. It allowed users to compose a single post and distribute it across multiple channels, with large language models automatically adapting the tone, format, and length for each platform's audience — an early and practical application of LLMs in a production workflow that I found genuinely exciting to build.

React.js
Next.js
Supabase
AWS
Vercel
Open AI
Stripe

Software Engineer

Try it on AI
Aug 2023 - Apr 2024

At Try it On AI, I stepped into a full-stack generalist role, taking ownership of both frontend features and backend integrations across the platform.

The core product used generative AI to produce studio-quality headshots from everyday photos. I built image editing tools, standardized client-side image resizing via the HTML Canvas element, and integrated external image processing and editing APIs to give users advanced editing capabilities — all powered by AI models running inference on the backend.

On the business side, I helped pivot the product toward B2B solutions: a corporate team feature for professional headshots, a commissioning program for photographers, and product linking to photos so users could "buy the look" they admired.

One of the more challenging features I developed was a public API dashboard that exposed the platform's AI-powered capabilities to external clients, enabling them to integrate headshot generation and image editing directly into their own products.

React.js
Next.js
Supabase
AWS
Vercel
Replicate
Open AI

Software Engineer

Deepwaters
Jun 2022 - Jun 2023

I joined Deepwaters as their primary front-end engineer and built the majority of their off-chain trading platform from scratch, alongside maintaining the company's marketing website.

This was the ideal environment for my strengths — precise product specs and the freedom to architect an entire web application from the ground up. I could invest in code architecture, component design, and overall quality rather than patching features onto an evolved legacy codebase.

Some of my favorite work here included the pie chart and price evolution chart — to me, charts sit at the intersection of data and geometry, making them some of the most complex and rewarding widgets to build. Both were implemented using only SVGs and primitive HTML elements, with no charting library involved.

Another highlight was the smart amount input: it auto-formats numbers in real time and prevents invalid states entirely. The core logic intercepts the "input" event after the user's action but before the UI updates, allowing seamless reformatting with zero flicker.

I also delivered the full responsive adaptation for mobile, reworking layouts, typography scales, and spacing across the platform.

React.js
GraphQL
Next.js
Ethers.js
Figma
Web3

Technical Lead

Minerva Gate
Mar 2022 - Dec 2023

At Minerva Gate, I led the delivery of multiple client-facing projects across different industries, alongside internal tools like the recruitment section of our website.

My role went well beyond code. I wore many hats: project manager, workflow designer, junior developer coach, and client liaison. In a small team, every skill gets used — and that breadth gave me a much sharper understanding of how startups operate end to end.

I learned how to plan efficiently at a small scale, how to research and adopt tools outside my core scope to improve the team's output, and how to keep multiple concurrent projects moving forward with limited resources.

As a stakeholder, my decisions carried real weight. It taught me that leading a startup demands more than hard work — it requires sharp judgment, the right insights at the right time, and the willingness to change direction when the evidence points that way.

React.js
Next.js
Figma
Django
Gatsby

Software Engineer Intern

Google
Sep - Nov 2016 & Jul - Sep 2017

Two internships at Google, both during university breaks. The first on the Google Cloud team in Warsaw, Poland; the second on Google Flights in Zurich, Switzerland. Both were front-end focused.

At Google Cloud, I built a standalone Chrome extension for the team — a developer tool for stats monitoring, quick actions, and web inspection. I also contributed a feature to the Kubernetes front-end using Angular.

At Google Flights, I implemented the price evolution chart for tracked flights. While I used Google Charts as the base, matching the design prototypes required creative workarounds — notably using CSS pseudo-elements to render custom tooltip tails that the charting API didn't natively support.

The biggest takeaway from Google was seeing how strong processes enable code to scale. Individual output feels smaller than at a startup, but the system ensures steady, reliable progress. I also learned that at scale, edge cases multiply — at Google Flights, many UI bugs stemmed from right-to-left language support alone.

Building and maintaining products at that scale is a discipline in itself, and team experience is the strongest predictor of success.

Angular
JSX
Protobuf

Software Engineer & Problem Setter

CS Academy
Jun 2015 - Nov 2017

I joined CS Academy straight out of high school with zero professional experience. We were a team of competitive programmers who loved algorithmic contests, and we built our own platform from scratch. It's still actively used today for contests, including Romania's national selection camps for international olympiads.

I prepared algorithmic tasks for the platform and built interactive widgets for several of them. The platform later partnered with IEEE to host their annual college-level 24-hour contests, and I served on the contest committee in 2017 and 2021.

Alongside the CTO, I contributed to Stem JS — a custom JSX front-end framework — where I built and styled foundational components: radio groups, tab areas, and a collapsible section divider.

A few standout front-end pieces I built here:

The workspace — a multi-language in-browser code editor built on Ace, wired to a backend for compilation and execution. I parsed build output to display inline errors and warnings directly in the editor.

The graph editor — an interactive visualizer where users can describe a graph via input or draw it manually. We started with D3, then rewrote it from scratch for full control. An essential tool for competitive programmers.

The diff tool — I implemented an O(N*D) diffing algorithm to find string differences, computed the relevant text blocks, highlighted them, and rendered gutter connections using SVG curves.

Stem.js
Algorithms
Data Structures
D3
Ace