Service 01
Full-stack development, from idea to production.
We are a full-stack development studio: custom web apps and SaaS, back offices, internal tools, and browser extensions, from the React frontend to the Python backend and PostgreSQL, deployed and run in production.
The problem
A product, not a handoff.
A product needs a backend, a frontend, a database, a cloud to run on, and someone to keep it working. Split across vendors, every handoff loses something, and too often the work stops at launch.
FundReachr started with no codebase, no team, and no infrastructure. Bonjout Beauty ran on gigantic Excel files. Bulq.ai's product was locked into a no-code frontend. We built what each one needed: the whole platform for FundReachr, an MRP for Bonjout Beauty, and for Bulq.ai the extraction platform and the backend that let it leave the no-code frontend.
What we build
Front to back, and after launch.
Web apps and SaaS
Customer-facing products with sign-in, billing, and the logic behind them: FundReachr's platform, with Auth0 sign-in, Stripe credit billing, and AI-written outreach, built end to end.
Custom MRP software and back offices
The tools a team runs the business with: Bonjout Beauty's MRP for purchasing, lot-level inventory, production, and COGS, FundReachr's admin back office, and Bulq.ai's app for validating extracted quotes.
Browser extensions
A Manifest V3 extension that fills investors' web forms with AI-written values, for FundReachr.
AI inside the product
LLM features where they remove manual work: investor messages written by AI in 11 languages, supplier quotes read by an LLM and checked by people, retailers' purchase orders parsed into the MRP. The data side is its own service: data collection and analysis.
Built to run
Automated tests, separate development and production environments, database migrations that run before each deploy, and the cloud infrastructure to keep it working.
How we work
Discover, design, build, run.
01
Discover
We scope the real problem, not the feature list. Business goals first, architecture second.
02
Design
System architecture and interface, designed together so the product holds up in production.
03
Build
Tight iterations, working software every week. You see progress, not status reports.
04
Run
We deploy on GCP, AWS, or Vercel, then monitor and evolve what we ship. You work directly with the people writing the code.
Stack
React, Python, PostgreSQL.
React on the web and Flutter on mobile; FastAPI, Flask, or Django on the backend; PostgreSQL at the core (Neon, Cloud SQL, or Supabase); deployed on GCP, AWS, or Vercel.
- React
- Flutter
- Python
- FastAPI
- Flask
- Django
- PostgreSQL
- Stripe
- Auth0
- three.js
Proof
From the database to the screen.
Fundraising · SaaS
FundReachr
An AI investor-outreach SaaS, built and shipped end to end
The whole SaaS: a Flask API with Auth0 sign-in and Stripe credit billing, an investor-data engine, email-delivery workers, four React apps, and a browser extension, live on Google Cloud.
Key figures
- Investor data synced into the app
- Daily
- Canonical record per investor
- 1
- Sectors in the investor taxonomy
- 43
- Languages for AI-written messages
- 11
Every investor gets a tailored email that the founder can preview, edit, regenerate or send, with delivery, open and click tracking. Demo data Beauty · E-commerce
Bonjout Beauty
From spreadsheets to an MRP platform
A full operations platform in Flask and React: purchasing, lot-level inventory, stock projection, production planning, landed-cost COGS, and Shopify sync, built on the database of the brand's previous internal app.
Key figures
- From Shopify order to the MRP
- Minutes
- Weeks of stock projection
- 12 to 52
- Backdated COGS restatement
- 2 years
- Sales data landed in BigQuery
- Daily
Purchase orders by vendor, with milestones and totals. Demo data Construction · AI
Bulq.ai
From no-code to a real engineering platform
The extraction platform behind Bulq.ai's data, which DataGem wrote nearly all of, and the foundation of the FastAPI backend-for-frontend that let customers leave the no-code app one at a time.
Key figures
- Differently formatted supplier quotes
- Thousands
- LLM spend in the re‑architected pipeline
- Capped
- Recovery sweep in the re‑architected pipeline
- 5 min
Every document's journey through extraction: status board, passes used and current stage. Demo data Software studio · Website
DataGem
This website, redesigned and rebuilt around the logo's own lines
This website: a design system, a WebGL engine that draws the logo's lines live, and React pages pre-rendered to static HTML, served by nginx on Cloud Run.
Key figures
- Lighthouse performance score on desktop
- 100
- Layout shift as pages load
- 0
- Compressed JavaScript to load a page, down from 440 kB
- 130 kB
A case study: the projects beside a sticky index that follows the reader, with hairline rows for goal, highlights and stack.
FAQ
Questions, answered.
What kind of software do you build?
Web apps and SaaS, back offices and internal tools, browser extensions, and the APIs and data behind them. FundReachr's investor-outreach SaaS, Bonjout Beauty's MRP, Bulq.ai's extraction platform, and this website are examples.
Do you build custom MRP software?
Yes. Bonjout Beauty's MRP covers purchasing, lot-level inventory, stock projection, production planning, landed-cost COGS, and Shopify sync. It is built on, and extends, the database of the brand's previous internal app.
Can you take over from a no-code tool?
Yes. For Bulq.ai we built the foundation of a FastAPI backend-for-frontend that let customers leave the no-code WeWeb app one at a time, without a big-bang migration.
Which technologies do you use?
React on the web, Flutter on mobile, FastAPI, Flask, or Django on the backend, and PostgreSQL at the core, deployed on GCP, AWS, or Vercel.
Do you stay after launch?
Yes. We deploy, then monitor and evolve what we ship: FundReachr's billing, authentication, and outreach were re-engineered as the product found its model.
Who will we work with?
Directly with the people writing the code, led by Pierre Denig. No account managers.