Services

Custom software, built to run itself.

We scope, build and operate automation — from a script that saves an afternoon every week to production platforms like SellerMagnet. Defined scope, fixed quote, working software.

Web scraping & data extraction

The data you need already exists — on marketplaces, portals and supplier sites. We build crawlers that collect it reliably, on schedule, in exactly the shape your systems expect.

  • Monitored pipelines that survive layout changes — failures alert us, fixes ship fast.
  • Structured delivery — REST API, CSV, database or straight into your warehouse.
  • Scheduling, retries and alerting built in from the first run, not bolted on later.
  • Responsible crawling — proxy management and rate limits that respect the target.
  • Legal & GDPR assessment per project — we tell you plainly what is feasible and what is not.
hive · crawler console
$ hive crawl --spider marketplace_de
· proxy pool healthy (24) · robots.txt honoured · 2 req/s
 crawled   40,231 pages · 0 failures · 3m 12s
 parsed    38,904 items · schema product_v3 valid
 warehouse.products   +38,904 rows

$ hive schedule --every 6h --alert slack
 scheduled · next run 06:00 UTC
· on failure: pages us, not you

Bots & process automation

The repetitive work between your tools — moving data, filing invoices, chasing status updates — is exactly what software is for. We automate it end to end.

  • Telegram, Discord and Slack bots your team actually talks to.
  • Browser automation for portals and legacy systems that have no API.
  • Back-office workflows — invoices, reports, data entry, reconciliation.
  • API-to-API glue that keeps systems in sync — on a schedule or reacting to events.
  • Alerts a human only when a human is actually needed.
hive · ops-bot (Telegram)
ops-bot · 07:00

Nightly import finished — 214 invoices processed, 3 flagged for review.

/flags

ops-bot · 07:04

#4801 VAT ID missing · #4805 duplicate · #4812 currency mismatch. All parked in Review.

/resolve 4805 duplicate

ops-bot · 07:05

✓ #4805 archived as duplicate — accounting export updated.

Custom scripts & tooling

Not everything needs a platform. Some problems need two hundred lines of well-written Python — delivered this week, documented, and yours.

  • One-off data jobs — migrations, merges, clean-ups, format conversions.
  • Excel and CSV pipelines that replace hours of weekly copy-paste.
  • Internal CLIs and small tools shaped around how your team works.
  • Days, not months — scoped tight, quoted fixed, shipped fast.
  • Documented and handed over — no lock-in by obscurity.
reconcile.py
# nightly reconciliation — scoped Mon, shipped Thu
from pathlib import Path
import pandas as pd

orders = pd.concat(map(pd.read_csv, Path("exports").glob("*.csv")))
crm    = pd.read_excel("crm_dump.xlsx", sheet_name="contacts")

clean = orders.merge(crm, on="customer_id", how="left")
clean.to_sql("orders_clean", engine, if_exists="replace")

print(f"{len(clean):,} rows reconciled")

Web apps & SaaS engineering

From first commit to paying customers. We build complete products — the same way we built SellerMagnet, our own Amazon seller platform, which we run in production today.

  • Python backends — Flask, FastAPI, PostgreSQL — built to be maintained for years.
  • Modern, fast frontends without framework sprawl.
  • Accounts, payments and billing wired in correctly the first time.
  • Deployment, CI/CD and observability from day one — not as an afterthought.
  • Proof, not promises — we operate our own product with the same stack.
See SellerMagnet live
GET /v1/status · 200 OK
{
  "service":     "sellermagnet-api",
  "status":      "operational",
  "version":     "2.14.3",
  "checks": {
    "api":       "ok",
    "workers":   "ok",
    "database":  "ok",
    "queue":     "ok"
  },
  "deployed_at": "2026-07-18T04:12:09Z"
}
Maintenance & retainers

Software needs an operator.

Scrapers break when targets change. APIs get deprecated. Requirements move. We stay on it — so what we build keeps producing.

Monitoring & alerting

Every pipeline we operate reports in. If a job fails or its output drifts, we know before you do — and usually fix it before you notice.

Fixed-response retainers

A monthly block of hours with a guaranteed response time. Changes, updates and fixes get handled — without a new quote for every request.

Continuous iteration

Automation compounds. We review what your systems produce and propose the next improvement — small, steady and measurable.

Stack

Tools we run in production.

Chosen for reliability and maintainability — boring where it should be, sharp where it counts.

Languages
Python TypeScript SQL
Frameworks
Flask FastAPI React Playwright Scrapy
Infrastructure
Docker AWS PostgreSQL Redis GitHub Actions
Process

Five steps. No surprises.

The same process for a three-day script and a year-long product: clear scope, visible progress, a working system at the end.

01

Scope call

Thirty minutes. You explain the problem, we ask the awkward questions and sketch the approach.

02

Fixed quote

A written scope, a fixed price and a delivery date — before we write a single line.

03

Build in weekly demos

Working software every week, not status slides. Course-correct while it is cheap.

04

Launch

Deployed to your infrastructure or ours — tested, documented and handed over properly.

05

Operate & iterate

Monitoring, maintenance and the next improvement — for as long as it earns its keep.

FAQ

Straight answers.

The questions every prospective client asks — answered the way we would answer them on a call.

How do you price projects?

Defined work gets a fixed-scope quote: one price, one timeline, in writing before we start. Ongoing operation — monitoring, maintenance, iteration — runs on a monthly retainer. No surprise invoices either way.

How fast can you deliver?

Scripts and small tools: days. Scrapers and bots: one to three weeks. Full products are scoped in milestones, so you see working software early — and keep seeing it every week.

Who owns the code and IP?

You do — in full, on final payment. Source code, documentation, infrastructure access. Nothing licensed back, nothing held hostage.

Is web scraping legal?

It depends on the target, the data and the jurisdiction — anyone giving you a flat "yes" is selling something. We evaluate robots.txt, terms of service and GDPR implications for every project, and when a request is not defensible we say so and propose an alternative.

Do you maintain what you build?

Yes. Every system we operate is monitored, and retainer clients get guaranteed response times — see maintenance & retainers above. If you prefer to run it yourself, we hand over documentation that makes that realistic.

Can you sign an NDA?

Gladly — and before any details are shared. Send us yours, or we will provide a standard mutual NDA.

Start a conversation

Have a project in mind?

Tell us what you want to scrape, automate or build — we reply within one business day.