English desktop home page of Mohsen's evidence-led portfolio
1 / 3

Mohsen’s evidence-led portfolio

A bilingual portfolio that helps technical and non-technical evaluators move from a relevant promise to inspectable evidence and a focused next step.

Role
Independent product designer and front-end engineer
Outcome
The repository now delivers a bilingual, responsive portfolio with audience-specific paths, verified public cases, deterministic decision support, focused contact entry, three registered self-case screens, and inspectable source.
Status
Delivered bilingual product and public source

This is an independent self-case, not client work. The screens and source verify delivery; no conversion, adoption, commercial, recruiter, or hiring result is claimed.

Project access

Outcome

Context and why it mattered

A polished surface without claim boundaries would make evaluation faster but less trustworthy; a repository dump would be truthful but would force every visitor to reconstruct relevance alone.

Outcome

Problem

A single generic portfolio pitch could not answer the different questions of a buyer, recruiter, delivery partner, enterprise evaluator, and technical reviewer without mixing claims, proof, and calls to action.

Delivery detail

Constraints

Every public claim and project visual had to remain traceable to a publishable source, while Persian RTL, English LTR, mobile constraints, motion preferences, and different reading depths shared one codebase.

  • Every public claim and project visual had to remain traceable to a publishable source, while Persian RTL, English LTR, mobile constraints, motion preferences, and different reading depths shared one codebase.

Delivery detail

Role and scope

The public repository and registered screens support independent authorship of the delivered product. No client, collaborator, traffic, conversion, or hiring claim is inferred.

Responsibilities

  • Audience-specific Build, Hire, Partner, and Enterprise routes
  • Evidence Bank, verified-project allowlist, and typed case narratives
  • Deterministic Navigator and three bounded Decision Lab tools
  • Persian and English responsive presentation with RTL/LTR behavior
  • Intent-preserving contact entry points and public implementation source

Included in scope

  • Audience-specific Build, Hire, Partner, and Enterprise routes
  • Evidence Bank, verified-project allowlist, and typed case narratives
  • Deterministic Navigator and three bounded Decision Lab tools
  • Persian and English responsive presentation with RTL/LTR behavior
  • Intent-preserving contact entry points and public implementation source

Outside scope

  • This is an independent self-case, not evidence of a client engagement or team-delivery claim.
  • The registered screens prove delivered surfaces; they do not establish visitor comprehension, conversion, adoption, or commercial impact.
  • The Navigator and Decision Lab produce bounded planning guidance, not autonomous professional advice or submitted project commitments.

Delivery detail

Route the evaluation job

The home and audience registries separate Build, Hire, Partner, and Enterprise questions before asking one page to carry every promise.

  1. 1

    Route the evaluation job

    The home and audience registries separate Build, Hire, Partner, and Enterprise questions before asking one page to carry every promise.

    • portfolio-audience-map
    • portfolio-home-en

Step outputs

  • portfolio-audience-map
  • portfolio-home-en

Delivery detail

Bind claims to publishable evidence

The Evidence Bank registers sources and visual assets, while the positive project allowlist prevents an incomplete or synthetic record from becoming a public case.

  1. 1

    Bind claims to publishable evidence

    The Evidence Bank registers sources and visual assets, while the positive project allowlist prevents an incomplete or synthetic record from becoming a public case.

    • portfolio-evidence-bank
    • portfolio-project-policy
    • portfolio-work-en

Step outputs

  • portfolio-evidence-bank
  • portfolio-project-policy
  • portfolio-work-en

Delivery detail

Keep assistance deterministic and bounded

Navigator retrieval and the three Decision Lab tools use typed catalogs, explicit inputs, and visible outputs instead of an open-ended chatbot that could invent portfolio claims.

  1. 1

    Keep assistance deterministic and bounded

    Navigator retrieval and the three Decision Lab tools use typed catalogs, explicit inputs, and visible outputs instead of an open-ended chatbot that could invent portfolio claims.

    • portfolio-navigator
    • portfolio-decision-lab

Step outputs

  • portfolio-navigator
  • portfolio-decision-lab

Delivery detail

Deliver one responsive bilingual system

Locale-aware routes, logical layout behavior, restrained cosmic enhancement, registered screens, and public source make the final product reviewable across language and device contexts.

  1. 1

    Deliver one responsive bilingual system

    Locale-aware routes, logical layout behavior, restrained cosmic enhancement, registered screens, and public source make the final product reviewable across language and device contexts.

    • portfolio-home-fa
    • portfolio-cosmic-contract
    • portfolio-repository

Step outputs

  • portfolio-home-fa
  • portfolio-cosmic-contract
  • portfolio-repository

Delivery detail

English desktop home

Shows the delivered top-level promise, proof sequence, and audience routing on a real product surface.

English desktop home

Shows the delivered top-level promise, proof sequence, and audience routing on a real product surface.

English desktop home page of Mohsen's portfolio
Contains visual content

Delivery detail

Persian mobile home

Makes the Persian RTL and narrow-screen delivery inspectable instead of treating localization as a text-only claim.

Persian mobile home

Makes the Persian RTL and narrow-screen delivery inspectable instead of treating localization as a text-only claim.

Persian mobile home page of Mohsen's portfolio
Contains visual content

Delivery detail

English case-study index

Shows the delivered evidence-led work index and need-based evaluation model.

English case-study index

Shows the delivered evidence-led work index and need-based evaluation model.

English desktop case-study index captured before the portfolio self-case was published
Contains visual content

Delivery detail

Public portfolio source

Lets a technical evaluator inspect the implemented routes, registries, tools, content boundaries, and responsive components.

Public portfolio source

Lets a technical evaluator inspect the implemented routes, registries, tools, content boundaries, and responsive components.

Contains visual content

Delivery detail

Typed audience journey map

Keeps each audience promise, objection, proof, and next step explicit and reviewable.

Typed audience journey map

Keeps each audience promise, objection, proof, and next step explicit and reviewable.

Contains visual content

Delivery detail

Evidence Bank

Separates verified evidence, outcome level, publishability, and asset usage from presentation copy.

Evidence Bank

Separates verified evidence, outcome level, publishability, and asset usage from presentation copy.

Contains visual content

Delivery detail

Positive publication policy

A project becomes public only after its slug is deliberately admitted to the verified allowlist.

Positive publication policy

A project becomes public only after its slug is deliberately admitted to the verified allowlist.

Contains visual content

Delivery detail

Deterministic Navigator flow

Routes a visitor through known intents and verified knowledge chunks without generating unsupported evidence.

Deterministic Navigator flow

Routes a visitor through known intents and verified knowledge chunks without generating unsupported evidence.

Contains visual content

Delivery detail

Three-tool Decision Lab

Exposes the Project Fit, MVP Cutline, and Architecture Trade-off methods as bounded, inspectable product flows.

Three-tool Decision Lab

Exposes the Project Fit, MVP Cutline, and Architecture Trade-off methods as bounded, inspectable product flows.

Contains visual content

Delivery detail

Static-first cosmic layer contract

Keeps the distinctive visual layer subordinate to content, motion preference, device capability, and one-canvas limits.

Static-first cosmic layer contract

Keeps the distinctive visual layer subordinate to content, motion preference, device capability, and one-canvas limits.

Contains visual content

Engineering detail

Architecture

Locale-aware routes read from typed audience, evidence, project, discovery, and tool registries. Pages admit only verified public content; the rule-based Navigator and Decision Lab use that same bounded context.

Visitor intent

other

Build, Hire, Partner, Enterprise, proof search, or technical review.

Locale-aware route shell

other

Persian RTL and English LTR pages share the same typed content contracts.

Audience and journey registries

other

Select the promise, objection, proof route, and next action for the evaluation job.

Evidence and publication registries

other

Admit verified claims, project slugs, source references, and proof assets.

Navigator and Decision Lab

other

Transform explicit visitor input into bounded guidance and an inspectable handoff.

Home, Work, case, and contact surfaces

other

Render responsive bilingual evaluation paths and preserve the selected intent into the next step.

  • visitorlocale-routes · enters in a chosen language
  • locale-routesaudience-registry · resolves the evaluation job
  • audience-registryevidence-registry · requests supportable proof
  • evidence-registrypublic-surfaces · admits claims and media
  • audience-registrydecision-surfaces · offers a bounded planning path
  • decision-surfacespublic-surfaces · returns an editable next step

Engineering detail

Route by evaluation job, not technology

Situation
Different visitors need different evidence and vocabulary, while a single technology-first pitch makes relevance their problem.
Decision
Use a shared system with explicit Build, Hire, Partner, and Enterprise routes.
Reasoning
The shared evidence model avoids duplicate truth while each route can answer one entry question and preserve one next action.
Cost / trade-off
More content contracts and cross-route consistency work than one generic landing page.

Alternatives considered

  • One long homepage for every visitor
  • Separate sites for every audience
  • One shared product with audience-specific routes and evidence

Engineering detail

Make publication evidence-led and deny-by-default

Situation
Legacy records, generated decoration, incomplete projects, and persuasive copy cannot all carry the same proof status.
Decision
Use the Evidence Bank, positive project allowlist, and narrative validation before rendering a public case.
Reasoning
Evidence status, source, asset usage, outcome level, and publication policy stay reviewable outside the presentation layer.
Cost / trade-off
Publishing a new case requires deliberate registration across the canonical typed registries.

Alternatives considered

  • Render every database project
  • Rely on page copy to explain uncertainty
  • Register evidence and admit public project slugs explicitly

Engineering detail

Prefer deterministic guidance to a generic chatbot

Situation
A conversational entry point can reduce ambiguity, but an open-ended model could invent capabilities, evidence, fit, or project commitments.
Decision
Use known intents, verified knowledge retrieval, explicit questions, typed calculators, and editable outputs.
Reasoning
The method remains visible and the output can preserve uncertainty instead of pretending to know missing context.
Cost / trade-off
The system handles a bounded set of decisions and cannot answer arbitrary portfolio questions conversationally.

Alternatives considered

  • Open-ended portfolio chatbot
  • Static FAQ only
  • Deterministic Navigator plus bounded decision tools

Engineering detail

Keep the cosmic layer distinctive but static-first

Situation
A recognizable visual motif can improve identity, but repeated canvases and unconditional motion would compete with proof, reading, and mobile performance.
Decision
Use typed hero, section, and quiet levels; keep one hero canvas at most and fall back before mounting it for constrained preferences or devices.
Reasoning
The identity remains visible while content hierarchy and lower-cost rendering stay primary.
Cost / trade-off
The motif is intentionally quieter outside the hero and differs by capability.

Alternatives considered

  • Continuous animated canvas in every section
  • Remove the visual motif entirely
  • One hero enhancement with static section and quiet fallbacks

Outcome

Result

The repository now delivers a bilingual, responsive portfolio with audience-specific paths, verified public cases, deterministic decision support, focused contact entry, three registered self-case screens, and inspectable source.

Engineering detail

Reflection

The stronger product model is a sequence: orient the visitor, establish trust early, expose the relevant case, offer depth on demand, and preserve context into one next step.

Lessons

  • A portfolio becomes easier to trust when proof status is a data contract rather than a tone-of-voice promise.
  • Technical depth works better as an optional reading layer after problem, ownership, and delivered result are clear.
  • Bilingual delivery must be visible in routing, direction, layout, and real screens—not only in translated strings.

Next steps

  • Run moderated comprehension reviews with technical, semi-technical, and non-technical readers and record where promise, proof, or next step is misunderstood.
  • Replace any duplicated route copy only after the intent registry shows the same evaluation job.
  • Add measured outcome language only when a named source, period, method, and limitation can travel with the claim.

Published content

Claim boundary

This is an independent self-case, not client work. The screens and source verify delivery; no conversion, adoption, commercial, recruiter, or hiring result is claimed.

Published content

Next step

Start with your goal

Share the outcome you need; the next step will be scoped around it.

Mohsen
Mohsen
[email protected]

Tehran, Iran

نسخهٔ فارسی

© 2026 Mohsen.