Prepared by: Murty BVNS, with workflow refinement input from LittleBird | Status: Living document — updated as new channels are added
Core principle: One source. One notebook. Two language tracks. Zero redundant steps. Production happens once, ahead of time, so nothing waits on a platform sync.
1. Why This Workflow Exists
Every episode starts as a single reflection, conversation, or shared post. From that one source, the workflow forks into two fully-produced, fully-branded outputs — English (Morning Walk with Murty) and Telugu (Antaryamitho Nadaka) — without duplicating research, re-explaining context to AI tools, or losing attribution to the original source along the way.
Problems this workflow was built to solve:
1. Redundant NotebookLM notebooks — earlier workflow opened a second, source-less notebook to write blog posts, losing the ability to credit the original source.
2. Missing source attribution — gratitude to original creators was an afterthought, not structural.
3. Idle time waiting on YouTube's auto-sync — solved in this version by moving all production to the evening before, so the sync completes overnight instead of during your limited morning hours.
4. Metadata generated too early (in even earlier versions) — fixed by generating the Unified Metadata Package right after both audios exist, so it reflects the finished spoken nuance rather than raw notes.
2. Why This Schedule — Seasonal, Not Fixed
This two-session split assumes production happens the evening before, on your main PC. That's confirmed, and here's the real reason it holds: in summer heat, you can't work efficiently on your main home-office PC and shift to browsing-only on a low-end laptop in the air-conditioned bedroom — and screen size makes focused work difficult given your eye strain, so evening production isn't practical in peak summer. As temperatures drop (autumn through spring), you're back on your main PC in the drawing room, often working late — which is exactly what this evening-first design needs.
Variant worth knowing: on days with enough time and energy, the entire workflow — production through Blogger/Facebook publishing — can be completed in one evening sitting rather than split across two sessions. As evening sessions expand, this is becoming the norm rather than the exception. The two-session structure below remains the reliable fallback (production never blocked on blog-writing energy, blog-writing never blocked on YouTube's sync) for shorter evenings.
Native scheduling, not manual posting: both Spotify for Creators and Facebook support native scheduled publishing. Rather than publishing instantly at whatever hour the evening session ends, use each platform's built-in scheduler to time the release for when readers/listeners actually wake up. This applies to Steps 5–7 (Spotify/YouTube/Audio.com) and Steps 12–13 (Facebook) — generate and finalize everything in the evening, but let the platform hold and release it at the right hour.
Retired step — manual audio QC: earlier in this workflow's life, episodes were listened to on NotebookLM during the walk, with any discrepancies corrected afterward in Adobe Podcast Studio before publishing to Spotify. That manual correction step is fully retired — NotebookLM's Audio Overview is now trusted and published directly, with quality control happening upstream (better prompts, better source material) rather than downstream (manual editing).
Seasonal note: if summer heat returns and evening PC work becomes impractical again, this workflow's session split should revert to a schedule built around your actual available hours at that time — worth revisiting rather than assuming this version holds year-round.
3. The Two Sessions, At a Glance
SESSION 1 — EVENING (production & primary distribution)
────────────────────────────────────────────────────────
Notes loaded into ONE notebook
↓
Manual filename slug written (YYYY-MM-DD-topic-keyword) —
a working filename only, not the SEO Title
↓
All 4 generations QUEUED back-to-back in NotebookLM
(EN Audio → TE Audio → EN Video → TE Video)
while you work on something else
↓
As EN Audio Overview lands: Audacity pass (+4–5 dB boost,
normalize/amplify) → export MP3 named with slug → tag
Artist/Album → re-upload into notebook as a source
↓
As TE Audio Overview lands: Audacity pass (normalize/amplify,
no dB boost needed) → export MP3 named with slug → tag
Artist/Album → re-upload into notebook as a source
↓
EN + TE Video Overviews land → saved with slug filenames,
held for Session 2's Facebook upload (no Audacity pass needed)
↓
Unified Metadata Package generated — Prompt 5, only now that
BOTH processed audios exist as sources (still last on purpose:
reflects finished spoken nuance, not raw notes)
↓
Publish EN episode → Spotify (Unified Metadata Title, not the slug)
Publish TE episode → Spotify (Unified Metadata Title, not the slug)
Publish to Audio.com (using Unified Metadata paragraph)
↓
→ YouTube auto-syncs for both episodes overnight while you sleep.
SESSION 2 — EARLY MORNING (blogging & legacy archiving)
────────────────────────────────────────────────────────
EN blog post generated (from notes + EN audio only)
↓
TE blog post generated (from notes + TE audio only)
↓
YouTube embeds are already live (thanks to the overnight sync) —
embed both Spotify + YouTube using last night's Unified Metadata
↓
Publish to Morning Walk with Murty (Blogger)
Publish to Antaryamitho Nadaka (Blogger)
↓
Native upload of last night's EN/TE Video Overviews to both
Facebook Pages, same Unified Metadata as caption/hashtags
↓
YouTube Music links added as first comment on each Facebook post
4. Step-by-Step Workflow
Session 1 — Evening: Production & Primary Distribution
Step 1. Load the source (reflection, Facebook post, YouTube video/reel, littlebird conversation notes) into one NotebookLM notebook. Write a quick manual filename slug (YYYY-MM-DD-topic-keyword) — this is a working filename for archiving/searchability only, not the SEO Title. It takes ten seconds and needs no AI call.
Step 2. Queue all four NotebookLM generations back-to-back — English Audio Overview (Prompt 1), Telugu Audio Overview (Prompt 2), English Video Overview (Prompt 6), Telugu Video Overview (Prompt 7) — and let them run while working on something else. ⚠️ Before the Telugu Audio Overview and Telugu Video Overview generate, manually select Telugu from NotebookLM's output language dropdown, and for the Telugu Video Overview specifically, the Explainer format (Short/Cinematic are English-only regardless of the dropdown). These are structured settings the prompt text cannot force on its own.
Step 3. As the English Audio Overview lands: Audacity pass — boost volume ~4–5 dB, normalize/amplify — export as MP3 named with the Step 1 slug, tag Artist/Album (Title stays blank for now), then re-upload it into the notebook as a source. (Open item: the audible quality impact of this dB boost + normalization pass hasn't been formally verified yet — flagged for testing, not yet a settled rule.)
Step 4. As the Telugu Audio Overview lands: Audacity pass — normalize/amplify only, no dB boost needed — export as MP3 named with the Step 1 slug, tag Artist/Album, re-upload into the notebook as a source.
Step 5. As the English and Telugu Video Overviews land, save each with the Step 1 slug as filename and hold them for Session 2's Facebook upload. No Audacity pass needed for these.
Step 6. Generate the Unified Metadata Package → Prompt 5, only now that both processed audios exist as sources in the notebook. This produces one shared title, summary, SEO description, Blogger labels, Facebook hashtags, and CTA-with-attribution — deliberately last, so it reflects the finished spoken nuance rather than raw notes or the working slug.
Step 7. Publish the English episode to Spotify, using the Unified Metadata's title/description (not the working slug). YouTube auto-sync begins.
Step 8. Publish the Telugu episode to Spotify, same Unified Metadata. Its own YouTube auto-sync begins.
Step 9. Publish to Audio.com using the Unified Metadata's micro-paragraph.
End of Session 1. Notebook holds: notes + processed English audio + processed Telugu audio. Both Video Overviews are saved and waiting. Both YouTube syncs complete overnight.
Session 2 — Early Morning: Blogging & Legacy Archiving
Step 10. Generate the English blog post → Prompt 3. Uses only the original notes + English audio.
Step 11. Generate the Telugu blog post → Prompt 4. Uses only the original notes + Telugu audio.
Step 12. Embed both the Spotify player and the (now-live) YouTube video, apply last night's Unified Metadata title/description/labels, and publish to Morning Walk with Murty (Blogger). (Embed source is always regular YouTube, not YouTube Music — YouTube Music does not offer embed codes for tracks/episodes, confirmed by checking directly.)
Step 13. Same embed + metadata approach, publish to Antaryamitho Nadaka (Blogger).
Step 14. Upload last night's English Video Overview (from Step 5) natively to the “Morning Walk with Murty” Facebook Page via Meta Business Suite, using the Unified Metadata's CTA/attribution line as caption and its hashtags. Never download from YouTube; never a lengthy Clipchamp edit. No Spotify/YouTube link in the post body.
Step 15. Same approach, upload last night's Telugu Video Overview to the “Antaryamitho Nadaka” Facebook Page.
Step 16. Once each episode is live on YouTube Music, add its link as the first comment on the matching Facebook post (never in the post body) — YouTube Music link, not the Blogger post, since many of your audience prefer listening on YouTube's platform over reading.
5. The Seven Prompts
Prompt 1 — English Audio Overview
(Paste into NotebookLM's “Customize” field. ⚠️ Fill in the source bracket before every run.)
You are two host curators for the 'Project Antaryami.' You are currently recording the official 'Audio Analysis' for the podcast “Morning Walk With Murty.”
SOURCE: [insert source — creator name, platform, title/link]
CRITICAL MANDATE:
1. Your very first sentence MUST be: “Welcome to the Morning Walk With Murty digital stroll.”
2. Never use the phrase “Deep Dive.” If you feel the urge to say it, replace it with “Gurukul Session” or “Digital Stroll.”
3. Frame the content through the 'Digital Prism'—explain that we aren't just summarizing text, we are extracting the 'Legacy Code' of Project Antaryami.
4. Identify the speaker as the “Architect of Antaryami” and the “Mentor of the Antaryami VA Academy.”
5. Tone: One speaker is the 'Strategic Auditor' (Digital Migration focus) and the other is the 'Wisdom Seeker' (Gurukul 2.0 focus).
SOURCE ATTRIBUTION (weave into the hosts' natural back-and-forth, do not announce as a separate segment): Using the source noted above, shortly after the opening line, have one host name the source and credit it in a tone of genuine appreciation — as a fellow traveler's spark, not a citation (e.g., “Before we extract today's Legacy Code, a quiet thank-you to [name] over on [platform], whose [post/video/reflection] set this whole stroll in motion”). The other host briefly acknowledges it before moving into the Digital Prism framing. Do not repeat the credit again — let the extraction of the Legacy Code carry the rest of the episode.
Prompt 2 — Telugu Audio Overview
(⚠️ Fill in the source bracket before every run.)
మీరు 'ప్రాజెక్ట్ అంతర్యామి' ప్రతినిధులుగా, “అంతర్యామితో నడక” పోడ్కాస్ట్ కోసం ఆడియో విశ్లేషణ చేస్తున్నారు.
SOURCE: [మూలం పేరు — వ్యక్తి/ప్లాట్ఫారమ్/టైటిల్/లింక్]
MANDATORY INTRO: మొదటి వాక్యం తప్పనిసరిగా ఇలా ఉండాలి: “అంతర్యామితో డిజిటల్ నడకకు స్వాగతం.”
SOURCE GRATITUDE: పైన ఇచ్చిన మూలాన్ని ఉపయోగించి, పరిచయం తర్వాత వెంటనే, ఒక వ్యక్తి ఈ మూలాన్ని పేరుతో ప్రస్తావిస్తూ నిజమైన కృతజ్ఞతతో ధన్యవాదాలు తెలియజేయాలి (ఉదా: “ఈ చర్చ మొదలుపెట్టే ముందు, [పేరు] గారికి ఒక చిన్న ధన్యవాదం — వారి [పోస్ట్/వీడియో/ఆలోచన] ఈ నడకకు స్ఫూర్తినిచ్చింది”). రెండో వ్యక్తి క్లుప్తంగా స్పందించి చర్చను ముందుకు తీసుకెళ్లాలి. ఈ ప్రస్తావన ఒక్కసారి మాత్రమే — మళ్ళీ పునరావృతం చేయవద్దు.
Vocabulary (Key Fixes): 'Information' బదులు సమాచారం; 'Purpose/Goal' బదులు ఉద్దేశ్యం/లక్ష్యం; 'Experience' బదులు అనుభవం; 'Future' బదులు భవిష్యత్తు; 'Deep Dive' బదులు సమగ్ర విశ్లేషణ/గురుకుల చర్చ; 'Digital Stroll' బదులు డిజిటల్ నడక.
Hybrid Tone: టెక్నికల్ పదాలు (AI, Internet, Data) ఆంగ్లంలోనే ఉంచి, అర్థాన్ని సులభమైన తెలుగులో వివరించండి.
Natural Telugu Mandate: భాష వాడుక భాషలో, సహజంగా ఉండాలి — పుస్తక భాషలా కాకుండా, ఎక్కువ సంస్కృత పదాలు వాడకుండా, సామాన్య శ్రోతకు అర్థమయ్యే రీతిలో, చిన్న వాక్యాలతో సంభాషణాత్మకంగా మాట్లాడాలి.
Identity: మూలాల్లో ఉన్న వ్యక్తిని “లీడ్ ఆర్కిటెక్ట్ మూర్తి BVNS” అని, “అంతర్యామి VA అకాడమీ మార్గదర్శి” అని గౌరవప్రదంగా సంబోధించండి.
Flow: ఇద్దరు వ్యక్తులు తమ అనుభవాలను పంచుకుంటున్నట్లుగా, జ్ఞానాన్ని అందిస్తున్నట్లుగా ఉండాలి.
Prompt 3 — English Blog Post
(Use in Session 2, from the notebook holding notes + English audio.)
Using the original source and the English audio overview in this notebook, write a blog post for “Morning Walk with Murty.” Do not restate the audio overview's structure — write fresh prose that expands on the ideas rather than summarizing the recording. Open with a one-line credit to the original source, naming the creator/platform and linking it if a URL is available. Where the post builds on a specific point from the source, attribute it inline (e.g., “As [name] pointed out...“). Close with a short, natural paragraph connecting the reflection to virtual work, digital entrepreneurship, or the Antaryami VA Academy mission, with a soft call-to-action pointing readers toward the Academy or VA Network — not a hard sales pitch, but a genuine “if this resonates, here's where it leads” bridge.
Prompt 4 — Telugu Blog Post
(Use in Session 2, from the notebook holding notes + Telugu audio.)
ఈ నోట్బుక్లో ఉన్న ఒరిజినల్ మూలం మరియు తెలుగు ఆడియో అనాలిసిస్ని ఉపయోగించి “అంతర్యామితో నడక” కోసం బ్లాగ్ పోస్ట్ రాయండి. ఆడియో నిర్మాణాన్ని పునరావృతం చేయవద్దు — ఆలోచనలను విస్తరించే కొత్త వచనం రాయండి. మొదటి వాక్యంలోనే మూలాన్ని (వ్యక్తి/ప్లాట్ఫారమ్) పేర్కొంటూ కృతజ్ఞత తెలియజేయండి, వీలైతే లింక్ చేర్చండి. వాడుక భాషలో, సహజంగా రాయండి — పుస్తక భాషలా కాకుండా. చివరిలో, ఈ ఆలోచనను అంతర్యామి VA అకాడమీ మిషన్తో సహజంగా అనుసంధానించే చిన్న పేరా రాయండి, మెల్లగా అకాడమీ వైపు నడిపించే విధంగా.
*(Run in Session 1, right after both audios exist. Produces ONE shared package used across both Blogger posts and both Facebook Pages. Confirmed: TE output is pure Telugu throughout — Summary, Description, Labels, Hashtags, CTA — except for genuinely untranslatable technical terms kept in English (same rule as the Telugu Audio Overview's vocabulary list). The Title alone keeps the clean two-part hybrid format for SEO reach.)*
You are the Executive Producer for 'Morning Walk With Murty' and 'Antaryamitho Nadaka.' Synthesize findings from ALL available audio and text sources in this notebook to generate a single 'Metadata Package' that reflects the final spoken nuance of both episodes.
SOURCE: [insert source — creator name, platform, title/link]
LANGUAGE MANDATE: “Unified” means one generation pass covering both language tracks — not one language reused everywhere. For every task below, produce two clearly labeled ver