Overall Score
Mrudula's portfolio is a sleek, technically impressive dark-themed showcase that nails design aesthetics and content depth but chokes on performance—with a 39-second LCP and a PageSpeed score of 58, this site needs serious optimization before it starts impressing actual recruiters.
The Roast
This portfolio is like a Ferrari with a bicycle chain: stunning to look at, technically sound under the hood, but absolutely wheezing when it tries to actually move. The irony of an AI/ML engineer building a production CV system that loads slower than a model training on CPU is *chef's kiss*—fix your Core Web Vitals before lecturing anyone about reliability and deployment. The 39-second LCP doesn't exactly scream 'production-ready inference pipelines,' does it?
🎯 Start Here
Google PageSpeed Insights
(Real metrics from Google)These scores come directly from Google's PageSpeed API. The AI scores above evaluate broader aspects like copy, trust signals, and conversion.
Core Web Vitals
Performance
SEO
Trust Signals
Conversion
Mobile
Accessibility
Copy & Messaging
Design & UX
Performance
This site is a performance disaster—39-second LCP is disqualifying for modern web standards, and the minification/unused CSS warnings indicate bloated deliverables that scream 'I built this locally and shipped it without optimization.' Even a student portfolio needs to load in under 3 seconds if you're claiming production-grade engineering chops.
Issues Found
- Largest Contentful Paint of 39.037s is a showstopper; likely caused by unoptimized hero image, render-blocking CSS, or lazy-load failures
- Minified CSS and JavaScript not served—suggests no build pipeline or production optimization; unused CSS still being loaded (indicates unused Tailwind classes or bloated frameworks)
- Mobile performance score of 58 indicates poor CWV metrics; TBT is 0ms (good) but LCP and CLS suggest image delivery and layout shift issues
Recommendations
-
Minify and Tree-Shake Assets high
Run production build with CSS/JS minification and purge unused styles (especially if using Tailwind or Bootstrap); deploy via CDN with gzip compression.
-
Optimize Hero Image Delivery high
Replace large unoptimized images with WebP format, use srcset for responsive variants, implement lazy loading on below-fold content, and preload critical LCP image with <link rel='preload'>.
-
Implement Caching Strategy high
Add cache-control headers (immutable for assets, 1-year for versioned bundles), enable Gzip on server, and consider Cloudflare or similar CDN to reduce time-to-first-byte.
SEO
Meta tags and H1 are well-crafted, and structural markup (nav/main/footer landmarks, skip link) is solid, but PageSpeed SEO score of 0 signals zero schema markup and missing critical on-page optimization signals that would help this portfolio rank.
Issues Found
- Zero schema markup detected—no JSON-LD for LocalBusiness, Person, or structured Project data; Google can't parse the case studies, skills, or work history as first-class entities
- Missing keywords in headings: H2s like 'Engineering Background' and 'Featured Case Studies' don't target search intent (e.g., 'Computer Vision Projects' or 'AI Model Deployment Case Studies')
- No internal link anchor text optimization—links like 'View Architecture' and 'Full Case Study Coming Soon' are generic and don't convey topic relevance for crawlers
Recommendations
-
Implement Structured Data (JSON-LD) high
Add Person schema for Mrudula, Project schemas for case studies (with tags for tech stack, outcomes), and OrganizationalAffiliation for work experience to enable rich snippet eligibility.
-
Optimize Heading Copy for Keywords high
Rewrite H2/H3 copy to include high-intent keywords ('Computer Vision Case Studies,' 'AI Model Deployment & Optimization,' 'Production ML Systems Architecture') without sacrificing readability.
-
Anchor Text Optimization medium
Replace generic CTA anchors ('View Architecture', 'View GitHub Profile') with descriptive text that includes keywords ('YOLOv5 Rack Damage Detection System' or 'Explore 8 Open-Source ML Repositories').
Trust Signals
GitHub proof-of-work (8 repos, 1 follower, Aug 2026 activity—note: future date suggests test data) and case study credibility are present, but lacking are: testimonials, social media follower counts, credentials/certifications, client logos, press mentions, or specific company names that would anchor trust for recruiters.
Issues Found
- GitHub stats show 8 public repos and only 1 follower—while transparency is good, single-digit followers on a 'production ML engineer' portfolio undercuts credibility; follower count feels low for someone claiming shipped systems
- No client testimonials, case study client logos, or company endorsements visible—'Shipped Industrial CV Pipelines @ RAMS Digital' lacks third-party validation or client logo to establish real-world trust
- Missing external validation signals: No LinkedIn recommendations visible, no awards/certifications displayed, no press/blog mentions, no speaking engagements—portfolio relies entirely on self-reported work
Recommendations
-
Add Testimonials or LinkedIn Recommendations high
Embed 2-3 short testimonials from colleagues, managers, or clients (e.g., 'Mrudula shipped a CV system 3x faster than expected' — Manager, RAMS Digital) with name/title to add third-party credibility.
-
Showcase Credentials & Achievements medium
Add a section for certifications (e.g., 'TensorFlow Certified'), hackathon wins, research publications, or speaking engagements to anchor expertise beyond portfolio projects.
-
Display Client/Partner Logos (if available) medium
Include logos for RAMS Digital or other companies featured in case studies to signal real-world adoption and client-facing credibility.
Conversion
Multiple CTAs exist ('View Architecture', GitHub, LinkedIn, Resume, 'Open to AI/ML Roles') but lack hierarchy, urgency, or clear win-to-action mapping; the site optimizes for awareness over conversion, missing opportunities to drive recruiter engagement or email signups.
Issues Found
- 'Open to AI/ML Roles' badge appears twice (hero + side panel) but doesn't link anywhere—should route to contact form or email; currently just static text signaling availability without conversion mechanism
- No contact form visible on desktop or mobile (despite 'Contact' in nav) or email link in hero—conversions likely route to external links (LinkedIn, GitHub) rather than owned channels
- CTA hierarchy is flat: 'View Architecture', 'GitHub', 'LinkedIn', 'Resume' all styled similarly with no visual priority; unclear which action recruiter should take first
Recommendations
-
Create Primary CTA with Urgency high
Make 'Open to AI/ML Roles' a clickable, prominent button in the hero that opens a contact modal or email form; style it distinctly (e.g., filled button vs. outline) with copy like 'Get In Touch / Let's Talk' to drive direct conversions.
-
Implement Contact Form or Email Link high
Add a simple contact form (name, email, message) accessible via the Contact nav link, or place a prominent 'Email me' link (mailto: in hero) to capture recruiter inquiries directly.
-
Establish CTA Hierarchy medium
Use visual distinction (filled primary CTA, outline secondary, plain text tertiary) to guide recruiter attention: Primary = 'Contact/Apply', Secondary = GitHub/LinkedIn, Tertiary = Resume/Architecture.
Mobile
Responsive layout adapts well (hamburger nav, stacked sections, no horizontal scroll), but dense vertical stacking, small touch targets, and tall viewport requirements (7014px total height) create friction—mobile feels like a long, cramped scroll rather than a guided experience.
Issues Found
- Skill tags and tech badges are small (likely <44px tap targets); on mobile, tapping individual tags is error-prone; should be larger or grouped into easily-tappable pill buttons
- Case study cards on mobile appear in single column with truncated descriptions ('Full Case Study Coming Soon')—reading engagement drops compared to desktop's side-by-side layout
- Hero section still uses center-aligned text and vertical stacking; on narrow viewports, the profile card and main content compete for attention rather than flowing naturally
Recommendations
-
Increase Touch Target Sizes high
Ensure all clickable elements (buttons, links, tag pills) meet 44x44px minimum; increase padding on skill tags and case study links to reduce mis-taps on mobile.
-
Optimize Case Study Preview on Mobile medium
Show full case study descriptions or expand-on-tap mechanism on mobile; avoid truncated text that requires scrolling to infer content.
-
Simplify Mobile Navigation low
Consider collapsing secondary sections (e.g., 'Beyond Engineering' bullet list) behind expand toggles to reduce initial scroll depth and improve perceived load time.
Accessibility
Solid foundation with 10 ARIA attributes, skip link, proper landmarks, and no missing form labels—but dark theme contrast ratios may fail WCAG AA on smaller text, and interactive elements like custom buttons and collapsible sections likely lack proper keyboard focus indicators.
Issues Found
- Dark navy/black backgrounds with light blue or cyan accent text—likely fails WCAG AA (4.5:1) contrast on body copy and skill tags; verified through contrast checker needed
- No visible keyboard focus indicators observed on buttons (e.g., 'View Architecture' CTA, 'GitHub'/'LinkedIn' links)—keyboard users won't know where they are
- Custom accordion-style sections ('01 Data', '02 Model', etc.) on mobile—unclear if these are keyboard navigable or have aria-expanded/aria-controls attributes
Recommendations
-
Audit & Fix Contrast Ratios high
Run WCAG contrast checker on all text/background combos; ensure at least 4.5:1 for normal text, 3:1 for larger text; lighten cyan accents or darken backgrounds as needed.
-
Add Visible Focus Indicators high
Apply :focus-visible styles to all interactive elements (buttons, links, form inputs) with 2-3px outline in a contrasting color (e.g., bright cyan or white) for keyboard navigation.
-
Verify Accordion Keyboard Support medium
Ensure collapsible sections support keyboard navigation (arrow keys, Enter/Space to toggle) and have proper aria-expanded, aria-controls, and role attributes for screen readers.
Copy & Messaging
Copy is technically precise and demonstrates deep expertise—'introspective uncertainty calibration,' 'edge latency,' 'human-in-the-loop auditability'—but reads like a research paper abstract rather than persuasive narrative, missing the chance to connect features to recruiter pain points.
Issues Found
- Value proposition buried: The first 100 words are jargon-heavy ('reliability-first,' 'edge deployment') without first stating *what problem* Mrudula solves for employers
- Case study descriptions are problem-engineering-result focused but lack quantified impact (e.g., 'Owned the dataset layer' doesn't tell a recruiter how much time/cost was saved)
- Call-to-action ('View Architecture') is passive and vague; not clear what recruiter gets by clicking vs. 'See How I Reduced Latency by 60%' or 'Explore Production ML System I Built'
Recommendations
-
Lead with Recruiter-Focused Hook high
Rewrite the first paragraph to lead with outcome ('Ships production CV systems 3x faster') before technical chops, then anchor with proof from case studies.
-
Add Quantified Results to Case Studies high
Update 'Shipped Industrial CV Pipelines' and case study descriptions with measurable wins (latency reduction %, accuracy lift, deployment time, cost savings) to make impact tangible.
-
Strengthen CTAs with Value medium
Replace vague CTAs like 'View Architecture' with action-oriented, benefit-driven text such as 'See How I Optimized YOLOv5 for Edge' or 'Review Full Technical Approach'.
Design & UX
Dark theme with neon accents, clean card-based layouts, and strong visual hierarchy create a modern, tech-forward aesthetic that actually matches the AI/ML positioning. The design is cohesive and professional, but the dense information architecture and lack of whitespace on mobile borders on cramped.
Issues Found
- Mobile layout stacks vertically with minimal breathing room—the 'HOW I ENGINEER AI' section and multi-step process feel claustrophobic on 375px screens
- Hero section uses center-aligned text which can feel less impactful than left-aligned alternatives; right-column card placement on desktop doesn't follow left-to-right reading patterns
- Inconsistent spacing between sections; the gap between 'Engineering Background' and 'Featured Case Studies' feels arbitrary
Recommendations
-
Improve Mobile Breathing Room medium
Increase vertical padding on mobile cards (particularly the engineering workflow and hero sections) and reduce font sizes slightly to prevent text density from overwhelming small screens.
-
Optimize Hero Layout for Attention low
Consider left-aligning hero headline text and repositioning the profile card to the right with intentional negative space to create stronger focal hierarchy.
Think you can beat 61/100?
Get your site brutally analyzed by the same AI. 8 scores, a punch list of fixes, 60 seconds.