netยทdevs
Work

Proof, not promises

Selected engagements. Each case follows the same honest format: context, problem, approach, measurable result.

Lokad AI Chat screenshot 1Lokad AI Chat screenshot 2
LokadSupply Chain SaaS

Lokad AI Chat

An AI assistant embedded inside the Lokad supply chain platform. The chat guides users through the platform, surfaces the right contacts, and โ€” in Envision mode โ€” generates, validates, and explains executable Envision scripts in real time. Conversation context is preserved across sessions so users pick up where they left off.

82% of AI-generated scripts run correctly on the first try. 94% by the second. ~70% reduction in time to first working script.

.NET 9Vanilla JSLLM IntegrationAzure Event Store

82%

Scripts correct on first run

94%

Scripts correct on second run

~70%

Faster time to working script

99%

Context preserved across sessions

View case โ†’

Vertex Transcribe Service screenshot 1
VertexEducation / Media

Vertex Transcribe Service

An AI-powered transcription pipeline built to handle millions of minutes of Hebrew, Aramaic, and English educational lectures. The service automatically detects audio vs. video, splits long recordings at silence points, runs parallel transcription jobs on Kubernetes, post-processes Hebrew script and diacritics, verifies religious source references, and outputs timed subtitles and HLS streams โ€” all without manual steps.

99% transcription accuracy. 300+ concurrent jobs. Up to 60% lower AI API costs through dynamic model selection. Built to scale to millions of minutes.

PythonFastAPIGoogle Vertex AIGeminiFFmpegKubernetesAWS S3

99%

Transcription accuracy

300+

Concurrent jobs

up to 60%

API cost reduction via model selection

99%

Subtitle timestamp accuracy

View case โ†’

Media Converter Service screenshot 1
Internal / Multi-projectMedia Infrastructure

Media Converter Service

A Kubernetes-based HLS media conversion pipeline driven by AWS SQS. A file lands on S3, a queue message fires, a Kubernetes job spins up, and the result โ€” multi-resolution video streams, multi-bitrate audio tiers, thumbnails, and master playlists โ€” is back on S3 with no manual steps. The service handles multiple audio tracks per video, auto-detects available GPU codecs, and scales worker nodes to zero when the queue is empty.

Fully automated pipeline from S3 upload to production-ready HLS. Zero idle compute cost between batches. Supports 5 video resolutions and 5 audio bitrate tiers per file.

PythonFFmpegKubernetesAWS SQSAWS S3HLS

5 (240pโ€“1080p)

Video resolutions per output

5 per track

Audio bitrate tiers

zero (scales to 0)

Idle compute cost

0

Manual steps in pipeline

View case โ†’

BGU Public Web Platform screenshot 1BGU Public Web Platform screenshot 2BGU Public Web Platform screenshot 3BGU Public Web Platform screenshot 4BGU Public Web Platform screenshot 5
Ben-Gurion University of the NegevHigher Education

BGU Public Web Platform

The central public digital platform for one of Israel's leading universities โ€” 80,000 pages, ~1,000 content editors, ~300 requests per second at peak. We led the architecture, built a custom Lucene-based search across 10+ specialized page types, implemented a rule engine for benefits eligibility logic, designed a per-subtree access control system for the editorial team, and completed a full infrastructure migration from Azure to AWS (ECS Fargate, RDS, S3) without downtime.

Stable at 300 RPS. 15,000+ staff profiles updated daily via incremental sync. Complex eligibility rules managed by editors without code changes.

C#.NETUmbraco CMSVue.jsAWSLuceneDocker

~80,000

Pages in production

~300 RPS

Peak load handled

~1,000

Content editors supported

15,000+

Staff profiles synced daily

View case โ†’

BGU Student Portal screenshot 1BGU Student Portal screenshot 2BGU Student Portal screenshot 3
Ben-Gurion University of the NegevHigher Education

BGU Student Portal

A closed, authenticated portal for students and faculty, running on shared infrastructure with the public site. Every content item is personalized across 8 criteria โ€” faculty, year, user type, and more. The portal integrates with 10+ APIs: schedules, grades, course payments, assignments, and Moodle SSO. Editors preview how any page will look for a specific user before publishing.

Fully personalized experience per user. Seamless Moodle SSO. Shared infrastructure keeps operational cost low.

C#.NETUmbraco CMSAngularAWSSSO

8

Personalization criteria

10+

API integrations

yes

Infrastructure shared with public site

View case โ†’

Disctopia screenshot 1Disctopia screenshot 2Disctopia screenshot 3Disctopia screenshot 4
DisctopiaMusic Streaming

Disctopia

A music streaming platform where artists publish and monetize content directly. The website and the streaming web app were two separate products using different auth systems and backend logic. We unified them: one Auth0 identity layer, one API for news, comments, and likes, and a CMS-driven event notification banner that marketing can manage without touching code.

Single authentication layer across both platforms. Shared API eliminated backend duplication. Event banners deployable by marketing in minutes.

.NETUmbraco CMSAuth0REST API

1 (Auth0)

Auth providers consolidated

yes

Backend duplication eliminated

yes

Event banner deployable without dev

View case โ†’