Author
Jonathan Burk
Co-founder & CTO, sipsip.ai
From Lockheed Martin to Gala Games, Jonathan has spent years building AI systems that handle what people shouldn't have to do manually. At Gala Games he served as AI Lead, building an agent-agnostic LLM harness that takes a JIRA ticket and produces a fully reviewed merge request — end to end. At sipsip.ai he architects the transcription pipeline, summarization layer, and knowledge infrastructure that makes the product feel effortless.
Writing
Articles by Jonathan Burk
21 articles
What Is a Knowledge Management System? A Technical Guide for 2026
A knowledge management system captures, organizes, and surfaces what your team knows. Here's the technical architecture behind modern KMS — and why AI changes the design.
AI Knowledge Management Tools in 2026: How AI Changes the Way Teams Learn
AI knowledge management tools do what legacy KMS can't: ingest audio and video, distill content automatically, and surface insights before you search. Here's the honest technical comparison.
The Best Obsidian and Notion Alternative in 2026: AI-First Knowledge Management
Obsidian and Notion are excellent tools — but neither captures audio, distills content automatically, or delivers daily briefs. Here's the technical case for what AI-first knowledge management actually looks like.
5 Best Speech-to-Text APIs in 2026 (Benchmarked by a Dev Team)
We benchmarked OpenAI Whisper, Deepgram, AssemblyAI, Rev AI, and Google STT on real podcast and meeting audio. Here's which API ships accurate transcripts — and at what cost.
How to Do an Online Background Check Using Public Sources — A Practical Guide
Court records, news archives, LinkedIn, and public databases are richer than most people realize. Here's the full manual OSINT checklist — and where AI synthesis takes over.
How AI Background Checks Work: Source Triangulation, Hallucination Reduction, and Dossier Synthesis
Why AI background checks produce lower hallucination than ChatGPT, how source triangulation works, and what happens between 'give me a name' and 'here's the cited dossier.'
How AI Transcribes Voice Recordings to Text: The ASR Pipeline Explained
What actually happens between pressing 'upload' and getting a transcript back? A breakdown of the ASR pipeline — preprocessing, model inference, diarization, and why accuracy varies.
Your Daily Brief, Now in Discord: How SipSip AI Delivers AI Summaries to Your Server
Connect your Discord account and the SipSip AI bot will deliver your personalized daily brief directly to you — same AI summary, now wherever your team already lives.
Lessons from the Claude Code Leak: Architecture, Product Direction, and Data Security
Claude Code leak lessons for builders: AI product architecture, source map security, execution-system design, and practical data security controls used by sipsip.ai.
How AI Podcast Summarizers Work: Whisper, Chunking, and the Full Pipeline
The full technical pipeline behind AI podcast summarization — from RSS feed ingestion and audio extraction to Whisper transcription, semantic chunking, and LLM summary generation.
Can AI Actually Watch a Video? How AI Video Analysis Works in 2026
AI can't watch videos the way humans do — but it can analyze them effectively. Here's the actual architecture behind AI video analysis: transcription, vision models, and LLM reasoning.
YouTube Video Summarizer API: How to Build AI Summaries from Any YouTube Video
A step-by-step technical guide to building a YouTube video summarizer using the YouTube Transcript API and an LLM. Includes working code, architecture decisions, and how sipsip.ai handles this in production.
Video to Text: How to Get a Transcript from Any Video File
Whether it's a YouTube video, an MP4 file, a screen recording, or a webinar — here's how to extract the full text transcript from any video, fast.
10 Best Uniscribe Alternatives for AI Transcription and Summarization (2026)
Uniscribe offers transcription, mind maps, and summaries in one tool. Here are 10 alternatives that do some or all of this better — depending on what you actually need.
Speech-to-Text Accuracy in 2026: WER Benchmarks, Model Comparisons, and What Actually Determines Quality
Word error rate benchmarks for Whisper large-v3 vs. Deepgram Nova-2, what drives accuracy variance across recording conditions, and how to consistently get above 95% on your audio.
Open-Source Video Transcribers: The Best Free Options in 2026
Whisper changed everything for open-source transcription. Here's an honest look at the best open-source video transcribers — and why most people end up wanting a hosted alternative.
The Best NotebookLM Alternatives in 2026 (Compared Honestly)
Google NotebookLM is impressive — but it has real limitations. We compared the top alternatives for AI summarization, document Q&A, and content intelligence across every format that matters.
How to Automatically Summarize Meeting Recordings With AI (MP3, MP4, M4A)
You don't need a live meeting bot to get AI meeting summaries. Upload any recording — MP3, MP4, or M4A — and get a clean transcript, action items, and key decisions in minutes.
How YouTube Transcript Generators Work: ASR vs Caption Retrieval (Technical Deep-Dive)
Two fundamentally different pipelines produce YouTube transcripts — and the choice between them determines accuracy. Here's exactly what happens under the hood.
How Researchers Extract and Archive YouTube Video Content: A Systematic Workflow
A researcher's workflow for extracting, archiving, and citing YouTube video content. From accurate transcripts to batch processing, citation formatting, and building a searchable research archive.
How We Built a Subtitle-First Transcription Pipeline
Most transcription tools download audio and run Whisper on everything. We built a smarter approach that's 10x faster for 80% of videos.