Founder @ WebTurnerAI · Backend Systems Engineer · Workflow Automation Builder
Snapshot · Preview · Overview · Philosophy · Engineering Focus · Founder · Impact · Systems · Architecture · Algorithms · Stack · Focus
| Perspective | What You See Here |
|---|---|
| For Recruiters | Production-ready backend/full-stack systems, role-based architecture, measurable workflow impact |
| For Founders | Execution-first product thinking, operational bottleneck solving, fast shipping with technical depth |
| For Clients | Practical automation systems that reduce coordination overhead and improve reporting visibility |
Real-time dashboard views from shipped operational systems
I build operational software that removes manual coordination from day-to-day execution.
My work sits at the intersection of backend systems, workflow automation, AI-assisted tooling, and full-stack product delivery. I prioritize software that produces measurable operational gains: faster cycle times, fewer manual handoffs, and clearer decision visibility.
I am currently pursuing an MSc in Computer Science while shipping production software through WebTurnerAI.
I approach products as operating systems for real teams.
- Start from an expensive manual workflow, not from a feature list.
- Design state and permissions before UI details.
- Keep deterministic logic for critical execution paths.
- Use AI as a bounded transformation layer, not as source of truth.
- Measure improvements with practical operational metrics.
This approach keeps systems reliable, explainable, and useful under production constraints.
| Domain | What I Build | Typical Outcome |
|---|---|---|
| Backend Systems | Auth models, role resolution, data access boundaries, API-connected logic | Reliable operational behavior under multi-role usage |
| Workflow Automation | Structured workflow engines, doc/export pipelines, action-generation flows | Reduced manual preparation and coordination overhead |
| Operational Dashboards | Real-time views for teams, managers, and admins | Faster decisions and clearer execution state |
| AI-Assisted Tooling | Genkit-based flows, typed schemas, controlled model integration | Practical AI gains without pipeline fragility |
| Full-Stack Product Engineering | End-to-end systems from UX to deployment | Production-ready systems with direct business value |
WebTurnerAI is my operating vehicle for building execution-focused software.
- Systems over noise.
- Workflow clarity over feature bloat.
- Shipping value over speculative architecture.
WebTurnerAI focuses on software for organizations that need:
- Better internal workflow orchestration.
- Visibility across active operations.
- Faster turnaround on repetitive process-heavy work.
These are direct system-level outcomes from active projects.
| Metric | Value | System |
|---|---|---|
| Workflow compression | 8h to 1h (87% reduction) | CMMI Navigator |
| Execution volume | 13,000+ operations in 2 hours | Aura AI Task Manager |
| Priority model depth | 5-factor weighted sorter | Aura AI Task Manager |
| Unified presentation architecture | 18 sections driven by one state model | CMMI Navigator |
| Role-based operation modes | Visitor, Student, Teacher, Admin | College Management App |
| Institutional content coverage | 20+ pages | College Management App |
| Algorithm consistency discipline | 40 consecutive days | LeetCode Daily |
All metrics are tied to named repositories and implementation artifacts.
Stack: TypeScript, Next.js, Firebase, Genkit, Gemini 2.5 Flash, Framer Motion, fl_chart
Operational problem
Team prioritization usually degrades into manual triage. Deadlines, dependencies, effort, and value compete in unstructured ways.
Engineering implementation
Aura AI uses a two-layer intelligence model:
- Deterministic ranking and planning engine.
- Generative AI assistant for natural-language task ingestion and support.
Why it matters
This architecture preserves execution reliability while still providing flexible AI interaction.
Recruiter signal
Demonstrates algorithm design, real-time state handling, and production-grade product UX.
Founder signal
Shows ability to convert a coordination problem into a scalable execution system.
Links
Stack: TypeScript, Next.js, Genkit, Tailwind CSS, ShadCN UI, docx, pptxgenjs
Operational problem
CMMI kickoff preparation is traditionally manual, repetitive, and time-expensive.
Engineering implementation
One typed CMMIData model drives:
- 18 interactive sections.
- 3 AI orchestration flows.
- Browser-side DOCX/PPTX export pipelines.
Why it matters
A single source of truth keeps UI, AI outputs, and exported artifacts synchronized.
Recruiter signal
Strong state modeling, typed AI pipeline integration, and practical automation architecture.
Founder signal
Directly converts consulting effort into reusable productized workflow infrastructure.
Links
- Repository: https://github.com/Aicodebyprince/pptautomation
Stack: Flutter, Dart, Firebase Auth, Cloud Firestore, Firebase Storage, SharedPreferences, fl_chart
Operational problem
Academic workflows are usually fragmented across separate systems for attendance, syllabus, events, and institutional content.
Engineering implementation
Unified role-based platform with:
- Auth-driven role routing.
- Real-time data sync across user types.
- Session lifecycle logic with controlled expiry.
Why it matters
It centralizes institutional operations while preserving role-specific views and access boundaries.
Recruiter signal
Proof of role-based system design, mobile product execution, and real-time backend integration.
Founder signal
Demonstrates ability to unify fragmented operations under one maintainable software surface.
Links
Stack: Next.js 15, TypeScript, Tailwind CSS, Framer Motion, Resend
Operational problem
Most portfolio surfaces either look polished but shallow, or deep but unreadable.
Engineering implementation
Narrative-driven site architecture with:
- Typed section and routing system.
- Motion system for hierarchy and flow.
- API-backed contact workflow and SEO structure.
Why it matters
It communicates founder positioning and engineering depth with product-level UX quality.
Recruiter signal
Clear communication architecture plus strong modern frontend implementation quality.
Founder signal
Shows product storytelling discipline and brand-to-technical coherence.
Links
- Repository: https://github.com/Aicodebyprince/prince-portfolio
- Live: https://princebuilds.vercel.app
Stack: React 18, TypeScript, Vite 6, Monaco Editor, Tailwind CSS
Operational problem
Complex interface behavior in editor-like products requires strong state architecture to stay usable.
Engineering implementation
IDE-style UI with tab management, explorer behaviors, resizable panels, and keyboard interaction system.
Why it matters
Demonstrates product-level interaction engineering and composable frontend systems thinking.
Recruiter signal
Strong evidence of advanced UI state management and interaction-system architecture.
Founder signal
Ability to build polished high-complexity interfaces with clear product direction.
Links
- Repository: https://github.com/Aicodebyprince/Codepilot
Stack: React, Supabase Auth, Supabase Database, Row-Level Security
Operational problem
Users need a secure and searchable system for fragmented personal operational data.
Engineering implementation
Vault cards, sticky notes, and categorized retrieval backed by auth-aware row-level access boundaries.
Why it matters
Shows practical access-control implementation and secure CRUD product architecture.
Recruiter signal
Demonstrates secure data design using auth and row-level access controls.
Founder signal
Builds trust-centered utility software with clear everyday operational value.
Links
- Repository: https://github.com/Aicodebyprince/Helpful-Vault
Stack: Python, pattern-based DSA practice
Operational problem
Interview prep and algorithm development often fail without consistency and documented reasoning.
Engineering implementation
40-day structured pattern progression with explicit complexity analysis.
Why it matters
Signals repeatable problem decomposition and algorithmic discipline.
Recruiter signal
Consistent complexity-aware coding discipline over a fixed timeline.
Founder signal
Reflects execution consistency and systems-thinking mindset under daily constraints.
Links
- Repository: https://github.com/Aicodebyprince/leetcode-daily
Landing, operations dashboard, and analytics views
Generated deck artifacts from AI-assisted kickoff workflow
Founder narrative and project storytelling surface
If you are scanning quickly: the top Live Product Preview section is the primary visual summary.
Across projects, I use a consistent operational architecture pattern.
flowchart LR
U[Client Applications\nFlutter or Next.js] --> A[Identity Layer\nFirebase Auth or Supabase Auth]
A --> R[Role Resolution\nVisitor, Student, Teacher, Admin, Team Roles]
R --> D[(Data Layer\nFirestore or Supabase)]
R --> S[Storage Layer\nFirebase Storage or Local Files]
D --> AL[Deterministic Logic\nRanking, Planning, Session Rules]
D --> G[Genkit Flow Layer]
G --> M[Gemini 2.5 Flash]
M --> G
AL --> O[Operational Outputs\nDashboards, Lists, Alerts]
G --> O
D --> E[Export Pipelines\nDOCX, PPTX, PDF]
- Single Source of Truth: shared canonical model for UI, workflow logic, and exports.
- Role-Aware Boundaries: access control is enforced from auth to data consumption.
- Real-Time Propagation: event and state updates push directly to active clients.
- Deterministic Core: critical planning logic remains algorithmic and testable.
- AI as Bounded Layer: AI transforms structured data, but does not own system state.
- Output Symmetry: the same data powers dashboards and exported artifacts.
| Decision | Tradeoff | Why Chosen |
|---|---|---|
| Serverless backend (Firebase/Supabase) | Less low-level infra control | High iteration speed for product workflows |
| Browser-side document generation | Heavy client operations on complex exports | Removes server bottlenecks for content export |
| Deterministic scoring + AI assistant | Increased architecture complexity | Preserves reliability while enabling flexible interaction |
| Role-specific dashboards over one generic dashboard | More UI surface area | Better clarity and lower cognitive load per user type |
score = (w1 * deadline_proximity)
+ (w2 * dependency_count)
+ (w3 * value_estimate)
+ (w4 * effort_inverse)
+ (w5 * blocker_penalty)
Why it exists: manual prioritization does not scale under high task volume.
Operational value: creates a deterministic execution queue per user/team context.
Given:
capacity C (available hours)
tasks T with value v and duration d
Goal:
maximize total value with sum(d) <= C
Approach:
sort by value density (v / d) and allocate greedily
Why it exists: daily planning overhead compounds across teams.
Operational value: compresses planning into an automated ranked schedule.
User Data -> generate-action-items.ts -> phase action plan
User Data -> presentation-generator.ts -> structured slide content
User Data -> theme-ai-assistance.ts -> design guidance
Why it exists: untyped model outputs increase integration fragility.
Operational value: schema-governed outputs improve reliability and maintainability.
on app launch:
read cached identity data
if session_age_days <= 5:
route to role dashboard
else:
clear session and route to login
Why it exists: institutional users need low friction with controlled session expiry.
Operational value: practical balance between usability and access hygiene.
| Layer | Technologies | Why This Layer Matters |
|---|---|---|
| Frontend | Next.js, React, Flutter, TypeScript, Tailwind CSS | Fast product iteration with typed, maintainable UI systems |
| Backend Services | Firebase Auth, Firestore, Supabase, REST APIs | Low-ops infrastructure with real-time capability |
| AI Orchestration | Google Genkit, Gemini 2.5 Flash | Structured AI integration with typed flow boundaries |
| Mobile | Flutter, Dart | Single codebase for cross-platform product delivery |
| Automation Outputs | docx, pptxgenjs | Business-ready artifacts generated from runtime data |
| Hosting and Delivery | Vercel, Firebase Hosting | Reliable deployment and production iteration speed |
I am currently building the next WebTurnerAI operational platform for consulting teams:
- Client engagement tracking from a single operational dashboard.
- Automated status report generation from live execution data.
- AI-assisted weekly summaries for manager/client visibility.
- Role-based access for consultants, managers, and clients.
- Export-ready outputs for reporting and stakeholder communication.
I am open to:
- Backend engineering roles.
- Workflow automation consulting.
- Product engineering partnerships.
- Technical founder collaborations.
Prince Sherathiya · Founder @ WebTurnerAI · Portfolio · LinkedIn · Email