The unglamorous services that run at 3am without anyone watching.

C# and .NET server applications for scraping data at scale, wiring AI models into real workflows, and processing video with FFmpeg. I ran a system doing all three in production for two years. Available hourly or on retainer.

Describe your problem See what I've built →

collector.service — production

// the shape of a typical build
[Scheduled] 03:00 → scrape sources resolved, rate-limited, retried
[Extract] 03:04 → parse unstructured HTML → typed records
[Enrich] 03:09 → llm classify, summarise, deduplicate
[Capture] 03:11 → ffmpeg record stream, transcode, thumbnail
[Store] 03:26 → db idempotent write, checkpoint
[Alert] 03:26 → notify only when something actually broke
✓ run complete — 26m, 0 errors, next run 24h

Capabilities

Three things, done properly.

This is a separate practice from my games work. Different clients, different problems, same standard of delivery.

Data collection

Web scraping at scale

Long-running services that collect data on a schedule and keep collecting when sites change, rate-limit, or go down. The hard part isn't the first scrape — it's the thousandth.

Scheduled & event-driven collection
Structured extraction from messy HTML
Retry, backoff & failure alerting
Deduplication & incremental sync
Export to DB, CSV, API or webhook

AI integration

LLMs wired into real workflows

Models are easy to call and hard to depend on. I build the layer around them — prompt handling, structured output, retries, cost control — so the output is something your system can actually trust.

API integration & prompt pipelines
Structured / JSON output enforcement
Classification, extraction, summarisation
Token & cost monitoring
Fallback handling when the model fails

Media processing

FFmpeg video pipelines

Recording live streams, transcoding, clipping, watermarking and thumbnailing — automated and running unattended. Built around FFmpeg, wrapped in a service that recovers from its own failures.

Live stream capture & recording
Transcoding & format conversion
Automated clipping & thumbnails
Watermarking & overlays
Queue-based batch processing

Case study

Two years, one system, still running.

EngagementLong-term contract Duration2 years RoleDeveloper within the client's team StatusIn production

A production Windows service for scheduled collection and live video capture

A single client needed data pulled from a rotating set of sources on a fixed schedule, and live video recorded and processed alongside it — continuously, unattended, without someone checking on it each morning.

I built and extended the .NET service that did both: scheduled scraping with structured extraction, an FFmpeg-driven capture and processing pipeline, and the recovery logic that kept it upright when a source changed shape or a stream dropped mid-recording. It ran for two years and is still running. I worked inside the client's team, with most of the implementation on my desk.

C# / .NET Windows Service Web scraping FFmpeg Scheduling SQL

Client name withheld under a confidentiality agreement. Happy to talk through the architecture in detail on a call — I just can't name who it was for.

Pricing

Hourly, or fixed once the scope is clear.

Small jobs are quoted fixed-price. Anything open-ended or ongoing runs hourly, invoiced against a logged breakdown. Paid via Wise or Payoneer internationally, bank transfer or UPI within India.

Engagement	Rate	Notes
Scraping & data collectionScheduled collectors, structured extraction, exports and sync.	$15/hr	Min. 5 hours
AI integration & video pipelinesLLM workflows, structured output, FFmpeg capture, transcoding and batch processing.	$20/hr	Min. 5 hours
Fixed-scope buildA defined script, service or pipeline with a written spec and an agreed deliverable.	Quoted	Quote in 24h
Monthly retainerReserved hours each month for maintenance, extensions and on-call fixes.	from $550/mo	40 hrs reserved
Rescue & auditAn existing scraper or pipeline that broke, or that nobody left behind understands.	$60 flat	Findings in 48h

Not sure whether your problem needs an hour or a month? Send the description and I'll tell you honestly — including if you'd be better served by an off-the-shelf tool than by paying me to build one.

Before you write

Questions I get asked first.

Is scraping legal?

It depends entirely on the source, the jurisdiction and what you do with the data. I build systems that respect robots.txt and rate limits by default, and I'll flag it if a target looks like it would put you at risk — but I'm a developer, not your lawyer, and anything sensitive is worth a proper legal opinion.

Do you work in anything other than .NET?

Yes. C# and .NET are where I'm strongest for long-running server work, but I also build in PHP, JavaScript and Node — my Gamify2Grow platform runs on a PHP backend. If your stack is already decided, tell me and I'll say plainly whether I'm the right fit.

What are your hours?

I'm in Chennai, UTC+5:30. That's a full working overlap with the UK and Europe in my afternoon, and with US East Coast in my evening. Average reply time across my client work is about an hour during the day.

Who owns the code?

You do, outright, on final payment. Full source, deployment notes and a handover walkthrough. I don't retain licences or hold anything back.

Get in touch

Tell me what's breaking, or what you need built.

Send a description of the problem and I'll come back with an approach and a number — usually the same day. If it's not something I should be doing, I'll tell you that too.

Email me the problem WhatsApp

sairam.vs@jcjustcoders.in +91 99403 69159 Chennai · UTC+5:30