90-Day Creating Ai Avatar with HeyGen AI project study plan
A comprehensive, step-by-step lecture and instruction guide designed as an AI Expert Masterclass Curriculum. Each daily unit breaks down technical execution, underlying principles, and exact action steps to master HeyGen and AI video production over 90 days.
Phase 1: Setup & Foundations (Days 1–20)
Day 1: Account Setup & Credentials Architecture
Lecture: Enterprise AI deployment begins with secure authentication protocols. HeyGen relies on cloud rendering nodes tied to user credentials; setting up proper SSO or secure email pipelines prevents rendering pipeline interruptions.
Step-by-Step Instructions:
Navigate to
app.heygen.com.Click Sign Up and select Continue with Google or enter an enterprise email address.
Verify the account via the confirmation link sent to your inbox.
Navigate to Account Settings > Security and enable Two-Factor Authentication (2FA).
Day 2: Dashboard Navigation & Asset Architecture
Lecture: Understanding the spatial distribution of tools in HeyGen’s UI is crucial for workflow efficiency. The platform separates assets into Studio Workspace, Avatar Library, Voice Lab, and Brand Kits.
Step-by-Step Instructions:
Log into your dashboard and locate the primary left sidebar.
Click on Avatars to inspect Instant, Studio, and Photo avatars.
Select Voices to access pre-built TTS models and custom voice uploads.
Open Templates to review predefined aspect ratio grids (16:9, 9:16).
Day 3: Workspace & Team Configuration
Lecture: Workspaces act as isolated sandbox environments. Properly structuring folder trees, user roles, and workspace limits ensures clean project management when scaling production.
Step-by-Step Instructions:
Click your profile avatar in the bottom-left corner and select Space Settings.
Rename your default space to
Primary AI Video Lab.Create sub-folders in your main dashboard:
01_Drafts,02_Renders, and03_Assets.Configure default export parameters to 1080p resolution.
Day 4: Hardware & Hardware-Interface Verification
Lecture: While HeyGen renders on cloud GPUs, high-fidelity avatar creation requires precise local hardware inputs. Webcams require minimum 1080p uncompressed capture, and microphones require cardioid polar patterns with low noise floors.
Step-by-Step Instructions:
Verify camera specs: Ensure 1080p/60fps or 4K/30fps capture capability.
Test audio input: Set mic gain so peak signal sits between -12dB and -6dB.
Ensure network throughput has at least 25 Mbps upload speed for rapid raw asset transfer.
Day 5: High-Resolution Source Image Selection
Lecture: Photo-to-avatar diffusion models rely on clear facial landmarks. Pupil occlusion, motion blur, or uneven shadows degrade landmark tracking and cause artifacting during frame generation.
Step-by-Step Instructions:
Select or capture a high-resolution portrait photo (minimum 2048x2048 pixels).
Ensure neutral lighting with no strong directional shadows on the face.
Frame the shot as a standard chest-up headshot with a clean, static background.
Upload the image to Avatars > Photo Avatar > Upload.
Day 6: Video Capture Protocol for Avatar Training
Lecture: Motion-capture training data for AI avatars requires static shoulders, consistent eye contact with the lens center, and natural facial gestures without extreme head rotations.
Step-by-Step Instructions:
Mount your camera at eye level using a tripod.
Record a 15-second 4K video speaking directly into the lens.
Keep head movement within a 15-degree cone; avoid covering your mouth or blinking excessively.
Export the raw file using H.264 encoding at a high bit-rate (at least 20 Mbps).
Day 7: Acoustic Sampling & Noise Floor Optimization
Lecture: Voice cloning engines utilize neural acoustic models (e.g., FastSpeech, WaveNet variants). Background ambient noise creates spectral artifacts in cloned audio, leading to metallic or robotic sound.
Step-by-Step Instructions:
Record 2–3 minutes of uninterrupted, continuous speech reading a varied phoneme text.
Use a lossless audio format (.wav, 48kHz, 24-bit).
Run a basic noise-gate pass in audio software (or HeyGen’s built-in cleaner) to eliminate room echo.
Save the file as
voice_sample_master.wav.
Day 8: Neural Voice Cloning Setup
Lecture: Neural voice cloning extracts pitch, timbre, accent, and cadence. Training requires explicit consent validation to prevent unauthorized deepfake audio generation.
Step-by-Step Instructions:
Go to Voices > Create Voice > Instant Voice Clone.
Upload
voice_sample_master.wav.Complete the voice consent verification by reading the on-screen prompt into your microphone.
Submit the dataset for neural processing.
Day 9: Acoustic Fidelity & Prosody Testing
Lecture: Synthetic voices often fail on edge cases such as acronyms, technical jargon, or emotional inflections. Testing synthetic output across phonetically balanced sentences validates acoustic stability.
Step-by-Step Instructions:
Open the created custom voice in the Voice Lab.
Input test text containing numbers, technical terms, and questions: "Welcome to the AI Lab. Is this 100% accurate?"
Generate the sample and audit for cadence roboticness, breath pauses, and sibilance.
Adjust pitch and speed parameters (+/- 5%) if necessary.
Day 10: Photo Avatar Animation Architecture
Lecture: Animating static imagery uses facial warping and lip-sync mapping models. Image resolution and clear mouth contours determine the stability of the generated animation.
Step-by-Step Instructions:
Go to Avatars > Photo Avatar.
Choose the photo uploaded on Day 5.
Select your custom cloned voice as the default audio driver.
Input a short 10-second script and click Generate to test base animation stability.
Day 11: Gesture Triggering Mechanics
Lecture: Human speech is naturally accompanied by non-verbal gestures. Strategic placement of body language markers prevents static "talking head" stiffness.
Step-by-Step Instructions:
Open a new Studio project and select a Video Avatar.
In the text box, select specific words or pause points.
Click Gestures in the editor toolbar.
Insert natural gestures (e.g., Wave, Point, Open Hands) at natural clause transitions.
Day 12: Micro-Expression Mapping
Lecture: Facial action coding system (FACS) parameters dictate subtle eye blinks, eyebrow raises, and smile intensities, directly impacting viewer trust and realism.
Step-by-Step Instructions:
Access the avatar expression settings panel in the script editor.
Set the baseline emotion to Friendly or Professional.
Add pauses (using
<break time="0.5s"/>or the UI pause tool) to allow natural eye-blink cycles.Preview the frame-by-frame rendering to ensure micro-expressions match script tone.
Day 13: Lip-Sync Latency & Accuracy Validation
Lecture: Lip-sync models map phonemes (sound units) to visemes (visual mouth shapes). Misalignments cause the visual "uncanny valley" effect.
Step-by-Step Instructions:
Write a script rich in plosives (P, B, T) and labiodentals (F, V): "Professional AI video platforms deliver performance."
Render a preview clip.
Frame-step through the preview video to verify mouth closure aligns perfectly with audio plosives.
Day 14: Spatial Environment & Chroma Keying
Lecture: Digital avatars must sit naturally within background contexts. Inconsistent lighting angles between the avatar capture and background plate break realism.
Step-by-Step Instructions:
Navigate to the Background section in the video editor canvas.
Choose between Transparent (Green Screen), solid color, or 3D office environments.
Select a neutral studio backdrop with realistic depth blur (bokeh).
Ensure background lighting vectors match your avatar's highlights.
Day 15: Lighting Contrast & Color Temperature Calibration
Lecture: Visual synthesis relies heavily on matching color temperature (Kelvin scale) and contrast ratios between synthetic foreground elements and background plates.
Step-by-Step Instructions:
Select your background canvas.
Adjust avatar color settings: fine-tune brightness, contrast, and saturation sliders.
Warm or cool the avatar tone (+/- 5% Kelvin bias) to align with background ambient light.
Day 16: Script Formatting & Phonetic Transcriptions
Lecture: Text-to-speech engines interpret literal text strings. Complex terms, foreign words, and numbers must be formatted phonetically to avoid mispronunciation.
Step-by-Step Instructions:
Draft a script in the text editor window.
Replace numerical values with full spelling (e.g., convert "$500" to "five hundred dollars").
Use phonetic spelling for complex terms (e.g., write "HeyGen" as "Hey-Gen").
Apply explicit pause duration tags between major paragraphs.
Day 17: Primary Video Synthesis
Lecture: Video synthesis combines script processing, voice generation, viseme mapping, and visual rendering into a final MP4 container.
Step-by-Step Instructions:
Assemble all elements into the main canvas: Avatar + Script + Cloned Voice + Background.
Perform a final preview scrub.
Click Submit / Generate Video at the top right corner.
Monitor cloud rendering queue progression.
Day 18: File Export Standards & Codecs
Lecture: Export settings dictate bandwidth efficiency and visual clarity. Choosing proper container formats, bitrates, and resolution avoids re-compression artifacts.
Step-by-Step Instructions:
Once processing finishes, click Download.
Select 1080p Original or 4K High Definition based on subscription tier.
Choose
.mp4format (H.264 codec with AAC audio stream).Save the output file to
02_Renders/day18_test_render.mp4.
Day 19: Visual Quality Audit Protocol
Lecture: Systematic Quality Assurance (QA) identifies edge artifacts, frame dropping, audio clipping, and viseme slips prior to publication.
Step-by-Step Instructions:
Play the exported video in a media player at 100% scale.
Perform an Audio-Visual Audit: Check for robotic voice distortion, lip desynchronization, and boundary artifacts around the hair/shoulders.
Log any detected rendering artifacts in a review tracking sheet.
Day 20: Avatar Fine-Tuning Iteration
Lecture: Refinement involves adjusting micro-parameters based on audit logs to build a stable visual identity for long-term production.
Step-by-Step Instructions:
Re-open the project draft.
Adjust avatar scale on the canvas to ensure natural rule-of-thirds positioning.
Tweak voice speed by -5% to improve articulation clarity.
Re-render and confirm performance enhancements.
Phase 2: Intermediate Development (Days 21–50)
Day 21: Conversational Script Engineering
Lecture: AI scriptwriting requires short, punchy sentence structures, active voice, and minimal subordinate clauses to sound natural when delivered by synthetic avatars.
Step-by-Step Instructions:
Open your script development document.
Write a 60-second script structured as: Hook (5s) -> Core Value (40s) -> Call to Action (15s).
Keep sentence length under 15 words per sentence.
Read the text aloud to ensure natural phrasing before pasting it into HeyGen.
Day 22: Automated Caption Generation
Lecture: Captions increase video engagement and accessibility across digital platforms. HeyGen integrates automated Whisper-based speech-to-text engines for precise subtitle timing.
Step-by-Step Instructions:
In the Studio timeline, click on the Captions icon in the toolbar.
Select Auto Captions and choose the primary language (e.g., English).
Generate captions and verify timestamp positioning against audio speech curves.
Day 23: On-Screen Dynamic Text Overlays
Lecture: Lower thirds, bullet points, and callouts keep viewers visually engaged, complementing avatar speech with visual anchors.
Step-by-Step Instructions:
Select the Text tool from the left sidebar.
Add a Header Text box for major key points and a Lower Third template for speaker identification.
Apply entrance animations (e.g., Fade In, Slide Right) synchronized with script timestamps.
Day 24: Multi-Scene Timeline Sequencing
Lecture: Complex videos require multiple distinct scenes to structure information logically and avoid visual fatigue.
Step-by-Step Instructions:
In the bottom timeline bar, click Add Scene (+) to create Scene 2 and Scene 3.
Assign different background assets or avatar placements to each scene.
Apply scene transitions (e.g., Dissolve, Wipe) between scene boundaries.
Day 25: Brand Persona Creation & Voice Tone
Lecture: Establishing a consistent corporate or personal persona requires matching visual styling, tone of voice, and delivery speed across all videos.
Step-by-Step Instructions:
Define persona attributes: Authoritative, Technical, Approachable.
Save a custom template containing specific avatar framing, font styles, and color palettes.
Lock this template for future persona-driven videos.
Day 26: Scripting for Technical Skills Showcase
Lecture: Demonstrating technical skills requires balancing explicit script narrative with timed visual media (e.g., screencasts, diagrams).
Step-by-Step Instructions:
Draft a script describing an AI pipeline or software workflow.
Insert script markers indicating exact timestamps where technical graphics should appear alongside the avatar.
Day 27: Picture-in-Picture Demo Integration
Lecture: Picture-in-picture (PiP) allows the AI avatar to commentate directly on software demos, presentation slides, or portfolio projects in real time.
Step-by-Step Instructions:
Upload a screen recording video asset to Uploads > Media.
Add the video as the main background canvas element.
Scale down your AI avatar and position them in the bottom-right corner (PiP layout).
Day 28: Interactive Outro & Call to Action (CTA)
Lecture: The final 5 seconds of a video drive conversion. Design elements must guide viewer attention to external resources, links, or contact info.
Step-by-Step Instructions:
Create a dedicated final scene.
Insert dynamic text elements featuring contact information, social links, and website URLs.
Add a dedicated CTA gesture trigger to the avatar (e.g., pointing towards the link box).
Day 29: Background Audio Layering & Mix Balances
Lecture: Background music adds depth, but unchecked audio levels mask speech frequencies. Audio mixing requires ducking music tracks relative to the voice track.
Step-by-Step Instructions:
Select Music from the asset library or upload a
.wavaudio track.Drag the track onto the timeline's audio channel.
Adjust background music volume to -18dB to -22dB relative to the 0dB speech baseline.
Day 30: Multi-Track Timeline Audio Synchronization
Lecture: Multi-track alignment ensures sound effects, background audio, and vocal tracks trigger precisely on specific video frames.
Step-by-Step Instructions:
Expand the timeline editor view.
Align transition sound effects (e.g., whoosh, clicks) precisely to text pop-ups and scene cuts.
Play back the timeline at 1x speed to verify mix balance and timing accuracy.
Day 31: Advanced Gesture Placement
Lecture: Over-gesturing creates unnatural visual pacing. Strategic placement on key emphasis words maximizes impact.
Step-by-Step Instructions:
Identify emphasis words in your script (e.g., "Crucial", "First", "Transform").
Insert precise point, open-palm, or emphasis gestures directly on those exact word tokens within the script editor interface.
Day 32: Expression Curve Modification
Lecture: Dynamic emotional shifts (e.g., transitioning from neutral problem statement to smiling resolution) improve viewer engagement.
Step-by-Step Instructions:
Split a single scene's script into two distinct blocks.
Set Block 1 emotion to Serious / Neutral for describing a problem.
Set Block 2 emotion to Excited / Happy for introducing the solution.
Day 33: Multi-Language Translation Architecture
Lecture: HeyGen’s translation engine leverages neural voice translation with automatic lip-sync adaptation, preserving original voice timbre across different target languages.
Step-by-Step Instructions:
Take a completed single-scene video project.
Select Video Translate from the dashboard.
Choose target language options and check Match Voice Timbre and Lip-Sync Adaptation.
Day 34: Spanish Localization & Accent Validation
Lecture: Localizing content into Spanish requires validating regional dialect choices (e.g., Castilian vs. Latin American) and syllable length differences.
Step-by-Step Instructions:
Generate Spanish localized output for your demo project.
Inspect translated transcript for proper regional syntax.
Verify that longer translated sentences fit naturally within the visual frame without awkward truncation.
Day 35: Mandarin Localization & Tonal Precision
Lecture: Mandarin is a tonal language. Ensuring synthetic speech accurately preserves pitch contours is essential for listener comprehension.
Step-by-Step Instructions:
Generate Mandarin localized output for your project.
Review audio stream against pitch accuracy.
Adjust speech speed parameters if tonal transitions sound rushed or compressed.
Day 36: Brand Kit Asset Management
Lecture: Centralizing brand assets (fonts, color hex codes, logos) ensures visual consistency across video production teams.
Step-by-Step Instructions:
Navigate to Brand Kit in the left sidebar.
Upload brand logo SVGs and custom fonts (.ttf or .woff).
Input primary, secondary, and accent color hex codes (e.g.,
#0055FF).
Day 37: Watermark & Logo Placement
Lecture: Strategic watermark placement protects intellectual property without obscuring key visual information or lower-third captions.
Step-by-Step Instructions:
Open a video project canvas.
Drag your logo from the Brand Kit onto the overlay layer.
Set opacity to 70%–80% and position it in the top-right corner.
Lock the aspect ratio and position to prevent accidental shifts.
Day 38: Color Palette Grid Integration
Lecture: Aligning text cards, shapes, and lower thirds with brand hex codes reinforces brand identity throughout the video timeline.
Step-by-Step Instructions:
Select text and geometric shape overlays in your scene.
Apply your primary brand hex code to headlines and secondary hex code to body text.
Check contrast ratio against the background to ensure compliance with accessibility guidelines (WCAG AA standard).
Day 39: Cross-Scene Visual Continuity
Lecture: Sudden jumps in avatar scale, lighting, or background positioning break production value across multi-scene projects.
Step-by-Step Instructions:
Copy visual transformation parameters (X/Y coordinates and Scale percentage) from Scene 1.
Apply exact transformation values to the avatar in Scene 2 and Scene 3.
Ensure background brightness remains consistent across all cuts.
Day 40: API Architecture & Endpoint Fundamentals
Lecture: HeyGen’s REST API enables programmatic video generation, allowing external software applications to automatically trigger video renders via HTTP requests.
Step-by-Step Instructions:
Navigate to Space Settings > API Access.
Generate a new API Key and copy it to a secure location (e.g., environment variable file).
Review API Documentation for the
POST /v2/video/generateendpoint.
Day 41: Programmatic Video Generation Setup
Lecture: Automating video workflows involves sending structured JSON payloads containing avatar IDs, voice IDs, and script text strings to the API endpoint.
Step-by-Step Instructions:
Construct a basic JSON payload specifying
avatar_id,voice_id, andinput_text.Send a POST request using cURL or Python
requestslibrary to the endpoint.Handle the HTTP response and capture the returned
video_id.
{
"video_inputs": [
{
"character": {
"type": "avatar",
"avatar_id": "Your_Avatar_ID",
"avatar_style": "normal"
},
"voice": {
"type": "text",
"voice_id": "Your_Voice_ID",
"input_text": "Automated video generation active."
}
}
]
}
Day 42: Batch Script Processing Operations
Lecture: Batch rendering allows creation of multiple video variants (e.g., personalized sales outreach or multi-name announcements) from a single dynamic template.
Step-by-Step Instructions:
Prepare a CSV file containing dynamic variable columns (e.g.,
Name,Company,Custom_Text).Import the CSV into HeyGen’s template batch processing tool or process via API script loop.
Validate batch generation queues in the processing dashboard.
Day 43: Interactive Q&A Video Formatting
Lecture: Structuring content as a dynamic Question & Answer creates an engaging educational format.
Step-by-Step Instructions:
Set up a two-character canvas or rapid scene-switch template.
Scene 1: Avatar asks a common technical question (Text graphic overlay displays question).
Scene 2: Avatar shifts perspective slightly and delivers the detailed technical answer.
Day 44: Custom Persona Fine-Tuning
Lecture: Reviewing mid-tier outputs allows identification of visual flaws, speech hesitation, or accent drift prior to compiling final portfolio assets.
Step-by-Step Instructions:
Review generated batch assets from Days 41–43.
Adjust baseline speech speed parameters (+/- 2%) and re-position avatar framing to eliminate subtle background alignment errors.
Day 45: Intermediate Portfolio Draft Assembly
Lecture: Combining individual modules into a coherent sequence demonstrates comprehensive control over avatar creation, voice cloning, and dynamic graphics.
Step-by-Step Instructions:
Create a master timeline project.
Sequence: Intro Scene -> Technical Skills Demo -> Multi-Language Showcase -> CTA Outro.
Add appropriate scene transition effects between modules.
Day 46: Peer Review & Feedback Mechanics
Lecture: Peer review uncovers subjective visual issues (e.g., uncanny facial expressions or poor audio balance) that automated checks miss.
Step-by-Step Instructions:
Export portfolio draft version
v0.5.Share preview link with peers or team members with an evaluation rubric focusing on: Audio Clarity, Visual Realism, and Pacing.
Day 47: Feedback Integration & Iteration
Lecture: Systematically applying feedback improves render quality and refines production standards.
Step-by-Step Instructions:
Log peer feedback into a structured revision sheet.
Modify timeline elements based on feedback: adjust audio levels, fix typos, and tweak gesture timings.
Re-render affected scenes.
Day 48: Accessible Captioning Standards
Lecture: Accessible captions must strictly conform to contrast ratios, display size, and read-time guidelines (WCAG standards).
Step-by-Step Instructions:
Enable closed caption styling options.
Set font to a legible sans-serif type (e.g., Arial, Inter), bold weight, high contrast background box (black background with white text).
Ensure no line exceeds 37 characters or covers lower-third graphical elements.
Day 49: Inclusive Visual Design Verification
Lecture: Ensuring video materials remain readable and accessible across diverse viewing environments and device sizes.
Step-by-Step Instructions:
Preview video renders on both desktop and mobile display viewports.
Verify text readability on low-brightness displays.
Confirm avatar posture and facial visibility across 9:16 and 16:9 aspect ratios.
Day 50: Intermediate Milestone Export
Lecture: Compiling and exporting an interim master asset establishes a benchmark for Phase 3 advanced customizations.
Step-by-Step Instructions:
Perform a final timeline sweep.
Render the project at maximum available resolution (4K / 1080p High Bitrate).
Export file as
Phase2_Master_Portfolio_v1.0.mp4and archive source assets.
Phase 3: Advanced Development & Finalization (Days 51–90)
Day 51: Custom Studio Avatar Generation
Lecture: Studio-grade avatars offer higher visual fidelity than instant web-cam clones. Creating one requires professional camera hardware, studio lighting grids, and dedicated training passes.
Step-by-Step Instructions:
Set up professional three-point lighting (Key, Fill, Backlight).
Record a 2-to-5 minute 4K high-bitrate video reading a continuous script.
Submit raw footage to HeyGen Studio Avatar Creation engine for deep-learning model creation.
Day 52: Wardrobe & Digital Outfit Adjustments
Lecture: Digital wardrobe swaps allow a single base avatar model to present in different contexts (e.g., business formal vs. smart casual).
Step-by-Step Instructions:
Open Studio Avatar options in the dashboard.
Select Outfit Customization (where available for supported avatars).
Generate and assign business formal attire for corporate presentations and casual attire for social media clips.
Day 53: Custom 3D Virtual Environment Integration
Lecture: Placing avatars in high-end 3D environments (e.g., virtual newsrooms or modern offices) requires correct foreshortening and drop shadow integration.
Step-by-Step Instructions:
Import high-resolution 3D virtual background plates or renders.
Apply a contact shadow layer under the avatar's feet or torso to anchor them into the scene realistically.
Apply slight Gaussian blur (1-3px) to the background plate to create camera depth-of-field.
Day 54: Complex Gesture Timeline Scrubbing
Lecture: Combining non-verbal micro-gestures with complex speech cadence requires frame-accurate timeline management.
Step-by-Step Instructions:
Zoom in on the advanced timeline editor down to individual frame intervals.
Precise-align eyebrow raises to tonal pitch spikes in the audio waveform.
Align hand gestures to start 2 frames before the corresponding spoken word for natural movement physics.
Day 55: Rendering Pipeline Optimization
Lecture: Optimizing project assets reduces cloud rendering times and minimizes credit usage across multi-scene projects.
Step-by-Step Instructions:
Compress uploaded background video files using H.264 prior to importing into HeyGen.
Remove unused tracks, hidden layers, and offline media assets from the timeline.
Standardize project framerate to 30fps.
Day 56: Render Engine Speed Testing
Lecture: Auditing execution times across render presets helps establish predictable production timelines for client or enterprise deliverables.
Step-by-Step Instructions:
Take a standard 30-second test scene.
Render the scene using standard processing vs. priority cloud queues.
Track time-to-render metrics and document processing benchmarks.
Day 57: Advanced Script Compression & Pacing
Lecture: Script compression maximizes information density while maintaining low cognitive load and clear voice model delivery.
Step-by-Step Instructions:
Edit a master presentation script to eliminate filler words, passive constructs, and redundant phrases.
Adjust speech output speed to 1.05x in the voice settings to create a crisp, professional pace without introducing distortion.
Day 58: Final Portfolio Persona Script Engineering
Lecture: The final portfolio script must clearly outline your end-to-end expertise in AI avatar creation, voice cloning, API automation, and visual design.
Step-by-Step Instructions:
Draft a comprehensive 90-second portfolio master script:
0-15s: Synthetic avatar self-introduction.
15-45s: Showcase of core capabilities (Voice cloning, multi-language, dynamic overlays).
45-75s: API automation and workflow architecture explanation.
75-90s: Concluding CTA and contact info.
Day 59: High-Resolution Brand Assets Integration
Lecture: Integrating high-DPI vector logos, crisp icon sets, and branded motion elements elevates video quality to broadcast standards.
Step-by-Step Instructions:
Export all brand assets as transparent vector-derived PNGs (at 300 DPI) or SVGs.
Place icons and lower-third graphic overlays into designated media tracks.
Ensure all graphics maintain consistent keyline padding and alignment grids.
Day 60: Animated Lower Thirds & Logo Bugs
Lecture: Motion adds visual interest. Subtle keyframed animation on logo bugs and lower thirds holds viewer attention.
Step-by-Step Instructions:
Select your logo overlay asset on the canvas.
Apply a subtle entrance animation (Motion Scale Up, duration 0.5s).
Configure a persistent lower-third banner with an animated entrance effect synchronized with your avatar's audio introduction.
Day 61: Master Persona Avatar Video Synthesis
Lecture: Synthesizing the final master video brings together high-fidelity avatar generation, precise script cadence, and custom voice models.
Step-by-Step Instructions:
Assemble all finalized assets onto the primary master timeline.
Run a full preview sweep, verifying lip-sync, gesture alignment, brand colors, and background mix.
Submit project for final full-resolution 4K video synthesis.
Day 62: Multi-Asset Portfolio Assembly
Lecture: A complete AI video portfolio includes primary video assets along with supplementary supporting materials (case studies, code samples, API scripts).
Step-by-Step Instructions:
Organize output video files into clean directory structures:
/Portfolio/Videos/.Prepare brief PDF or Markdown technical write-ups detailing the creation methodology for each video sample.
Day 63: YouTube Channel Optimization
Lecture: Optimizing video assets for YouTube delivery requires setting up proper encoding, aspect ratios, thumbnails, and descriptive metadata.
Step-by-Step Instructions:
Export your master video in 16:9 aspect ratio at 1080p/4K resolution.
Log into YouTube Studio and upload the master showcase video.
Create a custom high-contrast thumbnail containing a screenshot of your AI avatar and dynamic title text.
Day 64: LinkedIn Native Video Optimization
Lecture: LinkedIn’s native feed prefers square (1:1) or vertical/horizontal videos under 3 minutes with burned-in closed captions for silent autoplay.
Step-by-Step Instructions:
Re-frame or create a 16:9 / 1:1 version of your showcase video in HeyGen.
Ensure high-contrast burned-in subtitles are enabled.
Upload directly to LinkedIn as a native video post with short, structured post text.
Day 65: Web Embedding & Responsive Frame Setup
Lecture: Integrating synthetic video assets into personal websites requires responsive HTML5 video containers or iframe implementations.
Step-by-Step Instructions:
Host the output video on a fast content delivery network (CDN) or Vimeo/YouTube.
Grab the embed code or construct an HTML5
<video>tag withcontrols,playsinline, andpreload="metadata"attributes.Embed the video on your personal website or portfolio page.
Day 66: Video Analytics Framework Setup
Lecture: Tracking viewer retention curves, click-through rates, and play rates provides objective data to guide content iteration.
Step-by-Step Instructions:
Embed Google Analytics or custom platform tracking on your website portfolio page.
Set up event triggers for video interactions:
video_start,video_50_percent, andvideo_complete.
Day 67: Target Audience User Testing
Lecture: User testing uncovers retention drop-offs and identifies points where synthetic delivery may feel artificial or distracting.
Step-by-Step Instructions:
Share the portfolio page link with a pilot group of 5–10 test users.
Collect quantitative data on watch time and qualitative feedback on avatar voice realism and content clarity.
Day 68: Feedback Aggregation & Categorization
Lecture: Organizing feedback into structured categories helps separate high-priority visual/audio glitches from subjective preferences.
Step-by-Step Instructions:
Tabulate user testing results in a feedback matrix categorized by: Audio Quality, Visual Realism, Pacing, and Messaging Clarity.
Identify high-priority issues that impact overall quality.
Day 69: Directed Portfolio Refinement
Lecture: Executing targeted fixes based on user feedback resolves final production issues before public launch.
Step-by-Step Instructions:
Re-open HeyGen source projects corresponding to identified issues.
Re-time rushing audio tracks, fix subtitle typos, or adjust lighting balance.
Re-render affected scenes and update hosted files.
Day 70: Comprehensive Accessibility Audit
Lecture: A final accessibility sweep ensures full compliance with accessibility standards (WCAG 2.1 AA).
Step-by-Step Instructions:
Verify transcript completeness and closed-caption accuracy.
Test color contrast on all text overlays using an online WCAG contrast checker.
Ensure screen-reader compatibility for text elements on the web hosting page.
Day 71: Video Search Engine Optimization (VSEO)
Lecture: Video SEO ensures target search engines accurately index and rank your video assets.
Step-by-Step Instructions:
Write keyword-optimized video titles, descriptions, and tags focused on "AI Content Creator", "HeyGen AI Expert", and "Virtual Avatar Specialist".
Include timestamps in the video description to generate automatic search chapters.
Day 72: OpenGraph & Social Metadata Configuration
Lecture: Configuring social metadata ensures links display rich media previews (large images, correct titles, descriptions) when shared on social platforms.
Step-by-Step Instructions:
Add OpenGraph tags (
og:title,og:description,og:image,og:video) to your website portfolio header code.Test link previews using the LinkedIn Post Inspector or Twitter Card Validator.
<meta property="og:title" content="AI Content Creator Portfolio | HeyGen Showcase" />
<meta property="og:description" content="Explore my fully automated talking AI avatar, custom voice clones, and API video workflows." />
<meta property="og:image" content="https://yourwebsite.com/assets/avatar_preview.jpg" />
Day 73: Video XML Sitemap Generation
Lecture: Submitting a video XML sitemap directly informs search engine crawlers about the location, duration, and metadata of embedded video content.
Step-by-Step Instructions:
Generate a
video-sitemap.xmlfile referencing your portfolio page URL, video source URL, thumbnail URL, title, and description.Upload the sitemap file to your web server root directory and submit it via Google Search Console.
Day 74: Technical System Performance Audit
Lecture: Auditing web page speed ensures that embedded video assets load quickly without hurting page performance metrics (Core Web Vitals).
Step-by-Step Instructions:
Run Google PageSpeed Insights on your embedded portfolio webpage.
Set video embeds to lazy-load below the fold to protect initial page load speeds.
Verify smooth playback across mobile devices and desktop screens.
Day 75: Security & API Key Protocol Review
Lecture: Securing production pipelines prevents unauthorized use of API keys, voice models, and proprietary assets.
Step-by-Step Instructions:
Audit API key permissions in HeyGen dashboard; restrict key scope where possible.
Ensure API credentials are stored in secure environment variables, never hardcoded in front-end client scripts.
Verify voice consent permissions remain locked and protected.
Day 76: Portfolio Final Polish Pass
Lecture: Performing a final visual check ensures all assets, lower thirds, audio levels, and transitions meet professional broadcast standards.
Step-by-Step Instructions:
Review every master video asset sequentially end-to-end.
Ensure uniform loudness across all clips (-14 LUFS target for web distribution).
Verify logo alignment and branding elements across all scenes.
Day 77: Presentation Rehearsal Protocol
Lecture: Preparing to present your work requires mastering both technical implementation details and strategic business value explanations.
Step-by-Step Instructions:
Outline a live presentation script explaining your 90-day learning process, avatar deployment, and API integrations.
Rehearse live commentary alongside your AI avatar showcase video.
Day 78: Live System Demonstration Recording
Lecture: Recording a live screen-capture demo proving real-time control over HeyGen, API calls, and asset editing provides definitive proof of technical capability.
Step-by-Step Instructions:
Launch screen recording software (e.g., OBS Studio).
Record a live walkthrough demonstrating: HeyGen Dashboard -> Custom Voice Engine -> API Request Execution -> Final Output Rendering.
Save output as
live_system_demonstration.mp4.
Day 79: Portfolio Master Video Compilation
Lecture: The final showcase video combines your synthetic avatar, live workflow recordings, and portfolio samples into a single master asset.
Step-by-Step Instructions:
Import all finished clips into a timeline editor.
Sequence: Synthetic Avatar Intro -> Live API/Workflow Capture -> Multi-Language Demos -> Final CTA Outro.
Export master file:
AI_Avatar_Master_Showcase_Final.mp4.
Day 80: Final Public Deployment Launch
Lecture: Publicly publishing your portfolio establishes your presence as an expert in AI video, voice cloning, and synthetic media automation.
Step-by-Step Instructions:
Deploy updated code to your live personal website server.
Publish the master showcase video to YouTube, LinkedIn, and social media channels.
Verify all video links, embeds, and contact forms are fully operational.
Days 81–85: Audience Engagement & Iterative Refinement
Day 81: Distribution & Initial Monitoring
Lecture: Multi-channel distribution requires actively monitoring viewer interaction, retention charts, and comment sentiment across platforms.
Instructions: Post showcase videos to target channels; track key performance indicators (views, watch time percentage, retention drop-offs) every 12 hours.
Day 82: Feedback Processing & Tactical Revisions
Lecture: Viewing community comments provides valuable feedback on perceived realism and areas where the content can be improved.
Instructions: Log viewer comments regarding voice pitch, speed, or visuals. Draft updated script lines or gesture adjustments to address constructive feedback.
Day 83: Persona Brand Alignment Update
Lecture: Updating avatar styling, wardrobe, or delivery tone based on initial platform performance keeps your brand fresh and aligned with audience expectations.
Instructions: Modify baseline avatar attributes in HeyGen settings; update background settings or lighting templates to refine overall visual quality.
Day 84: Advanced Performance Analytics Audit
Lecture: Deep analysis of analytics data isolates precise drop-off timestamps, showing exactly where audience attention drops.
Instructions: Open YouTube and LinkedIn analytics dashboards. Identify timestamps with sharp audience drop-offs; analyze corresponding visual/script elements to inform future edits.
Day 85: Iterative Video Regeneration Protocol
Lecture: Using modular rendering pipelines lets you quickly update specific video scenes without rebuilding the entire project from scratch.
Instructions: Re-render specific scenes identified during analytics reviews. Replace outdated clips on host platforms with the improved versions.
Days 86–90: Final Showcase, Deployment & Documentation
Day 86: Portfolio Asset Packaging
Lecture: Packaging video assets, code repositories, and technical write-ups into a unified presentation creates an accessible, comprehensive portfolio.
Instructions: Assemble all final video links, API code samples, CSV template workflows, and case studies into a clean master directory or landing page.
Day 87: Quality Control Verification
Lecture: A final quality assurance run guarantees that every frame, audio track, and caption displays without errors across devices.
Instructions: Conduct a complete playback sweep of all portfolio assets across desktop, tablet, and mobile devices. Verify audio clarity and transition smoothness.
Day 88: Executive Master Presentation Capture
Lecture: Creating an executive-level walkthrough video demonstrates your ability to explain complex technical workflows clearly to stakeholders.
Instructions: Record a comprehensive 3-minute video presentation featuring your AI avatar introducing your technical architecture, API pipelines, and completed portfolio.
Day 89: Strategic Platform Distribution
Lecture: Maximizing outreach requires executing a synchronized distribution plan across all professional channels.
Instructions: Upload final master assets to primary hosting channels (YouTube, LinkedIn, Personal Site). Publish technical summaries on platforms like Medium or Dev.to linking to your portfolio.
Day 90: AI Workflow Documentation & Certification
Lecture: Documenting end-to-end operational procedures transforms trial-and-error experience into a repeatable, scalable AI video production system.
Instructions:
Write a comprehensive Standard Operating Procedure (SOP) covering: Raw Data Capture -> Voice Cloning -> HeyGen Video Synthesis -> API Automation -> Final Export.
Share the completed documentation publicly alongside your portfolio links as tangible proof of AI video mastery.
Comments
Post a Comment