Build a Professional AI Evaluation Website for Latent Lens Conversion-Centric B2B Website

PROJECT OVERVIEW

I need an experienced web developer to build a complete, premium, responsive website for Latent Lens, an early-stage AI evaluation and quality assurance company.

Our services help AI businesses assess model quality, test AI agents, identify hallucinations, evaluate safety, compare models and organise human feedback and training-data quality services.

Target customers include international AI startups, AI labs, SaaS companies, AI product developers and enterprise technology teams.

The objective is to establish credibility, explain our services clearly and generate qualified customer enquiries. This is a complete website project, not a homepage-only design.

Do not invent clients, testimonials, partnerships, certifications, case studies, revenue or performance results.

1. DESIGN REQUIREMENTS

Create a distinctive, premium B2B technology website using midnight navy, near-black, electric blue, cyan, white and cool grey.

Use professional typography, generous whitespace, strong visual hierarchy, subtle gradients, clean icons, technical diagrams and consistent spacing.

Avoid generic templates, excessive neon effects, meaningless AI graphics, fake dashboards and unnecessary animations.

Create a reusable design system. Use our official logo if supplied, otherwise use a temporary replaceable brand mark.

2. REQUIRED PAGES

Implement the following pages with descriptive URLs, unique content and appropriate metadata.

CORE PAGES

1. Homepage: /
2. About: /about/
3. How We Work: /process/
4. Industries: /industries/
5. Insights: /insights/
6. Contact: /contact/
7. Privacy: /privacy/
8. Terms: /terms/

AI EVALUATION & TESTING
9. Overview: /services/ai-evaluation-testing/
10. LLM Evaluation: /services/llm-evaluation/
11. AI Agent Testing: /services/ai-agent-testing/
12. Accuracy and Factuality: /services/accuracy-factuality/
13. Hallucination Detection: /services/hallucination-detection/
14. Instruction Following: /services/instruction-following/
15. Human Preference Evaluation: /services/human-preference-evaluation/
16. AI Safety Evaluation: /services/ai-safety-evaluation/
17. Response Quality Scoring: /services/response-quality-scoring/
18. Model Comparison: /services/model-comparison/
19. AI Red Teaming: /services/ai-red-teaming/
20. Multilingual Evaluation: /services/multilingual-evaluation/

AI TRAINING DATA & HUMAN FEEDBACK
21. Overview: /services/training-data-human-feedback/
22. Data Annotation: /services/data-annotation/
23. Human Feedback: /services/human-feedback/
24. Response Ranking: /services/response-ranking/
25. Dataset Quality Assurance: /services/dataset-quality-assurance/
26. Evaluation Dataset Development: /services/evaluation-datasets/

INDUSTRIES
27. AI Startups: /industries/ai-startups/
28. Enterprise AI: /industries/enterprise-ai/
29. AI SaaS: /industries/ai-saas/
30. Conversational AI: /industries/conversational-ai/
31. AI Agents: /industries/ai-agents/

TEN INSIGHTS ARTICLES
32. What Is AI Evaluation? A Practical Guide
33. How to Evaluate an LLM Before Launch
34. How to Test AI Agents Beyond Simple Accuracy
35. How to Detect and Measure AI Hallucinations
36. AI Benchmarking: What Teams Should Measure
37. Human Feedback in AI Evaluation
38. Building a Reliable AI Evaluation Dataset
39. Red Teaming for AI Systems
40. Evaluating AI Before Production Deployment
41. AI Quality Assurance: A Practical Framework

TRUST PAGE
42. Illustrative Sample Evaluation Report: /sample-report/

The contact page must also support project-discussion enquiries.

3. HOMEPAGE

Suggested headline:

“Know Where Your AI Fails. Understand Why. Improve With Evidence.”

Include:

* Professional navigation and hero section.
* Clear explanation of AI quality problems.
* Our two main service categories.
* Individual service summaries.
* Evaluation methodology.
* Illustrative evaluation criteria and findings.
* Target audiences and use cases.
* Sample report preview.
* Evidence-based recommendations.
* Useful FAQs.
* Strong calls to action.
* Professional footer.

Primary CTA: Discuss Your AI Project.

Secondary CTAs: Explore Our Services and View Sample Evaluation Report.

Make the homepage visually distinctive, easy to scan and commercially persuasive.

4. SERVICE PAGES AND CONTENT

Every service page must have substantial, technically accurate, distinct content explaining the service, problems addressed, methodology, use cases, potential deliverables, client requirements, limitations, related services and relevant FAQs.

Clearly distinguish AI evaluation from model training, human feedback from annotation, accuracy from factuality, and benchmarking from production monitoring.

Do not promise guaranteed improvements, zero hallucinations or completely safe AI. Explain that scope, evaluation criteria and available capabilities depend on each engagement.

The ten articles must provide useful information, address distinct search questions and link naturally to relevant services. No thin or duplicated content.

5. SAMPLE EVALUATION REPORT

Create a professional report containing the evaluation objective, system, scope, questions, test design, criteria, example test cases, hypothetical findings, evidence, limitations, recommendations and retesting considerations.

Clearly label it “ILLUSTRATIVE EXAMPLE ONLY”. Never present hypothetical findings as actual client results.

6. NAVIGATION AND CONTACT FORM

Implement functional desktop navigation, service dropdowns, mobile navigation, breadcrumbs, internal links, CTAs and consistent headers and footers.

Create a contact form with name, work email, company, role if useful, AI system type, service interest and project description.

Validate inputs and provide clear error and confirmation messages.

If submission is simulated, explicitly state that no enquiry was sent. Document the backend integration required for production. Do not invent email addresses or booking links.

7. SEO AND PERFORMANCE

Implement:

* Unique titles and meta descriptions.
* Descriptive URLs and canonical URLs.
* Logical heading hierarchy.
* Contextual internal links.
* XML sitemap and appropriate robots directives.
* Open Graph metadata.
* Appropriate image alt text.
* Accurate structured data where applicable.
* Mobile-friendly, crawlable content.
* Proper handling of missing pages.

Use the confirmed production domain for canonicals. Prevent staging pages from being indexed unintentionally.

Do not promise rankings or guaranteed traffic. Do not invent reviews, company credentials or business facts.

8. RESPONSIVENESS AND ACCESSIBILITY

Test at 320, 375, 430, 768, 1024 and 1440 pixels.

Check mobile menus, horizontal overflow, clipped content, buttons, cards, forms and footer layouts.

Aim for WCAG 2.2 AA accessibility where applicable. Use semantic HTML, keyboard-accessible controls, visible focus states, readable text and accessible forms.

9. TECHNICAL REQUIREMENTS

Review our existing website and supplied prototype before proposing the implementation.

Use a lightweight, maintainable stack with reusable components, optimised assets and straightforward deployment.

Provide complete source code, styles, scripts, assets and documentation. Avoid unnecessary dependencies and recurring costs.

Do not expose credentials or API keys in frontend code. Document hosting, dependencies, deployment and ongoing costs.

The website and relevant accounts must remain under my control.

10. TESTING

Before delivery:

* Visit every page.
* Check all internal links and navigation.
* Test form validation and submission feedback.
* Verify metadata, canonicals and sitemap entries.
* Test all six viewport widths.
* Check for broken routes, missing assets and JavaScript errors.
* Fix identified problems.
* Provide an honest audit report showing actual test results and remaining limitations.

11. DELIVERABLES

1. Complete working website on an agreed staging or production environment.
2. Direct preview URL that opens without installing software.
3. Complete source code and custom assets.
4. Page inventory and developer documentation.
5. SEO implementation.
6. Testing and audit report.
7. Deployment and ownership handover.
8. Disclosure of recurring costs.
9. Agreed post-delivery bug-fixing support.

A screenshot, static mockup or homepage-only website is unacceptable.

12. MILESTONES

1. Review the existing website and approve the design direction.
2. Build the homepage and shared design system for approval.
3. Implement service and industry pages.
4. Implement all ten articles, sample report and contact form.
5. Complete SEO, responsive optimisation and testing.
6. Deploy and transfer the source code and documentation.

Use milestone-based payments. Do not remove agreed pages or reduce scope without written approval.

13. QUESTIONS FOR YOUR BID

Please provide:

1. Two or three relevant B2B, AI or SaaS websites you have built, with working URLs.
2. Recommended technology stack and reasons.
3. Confirmation that all required pages and content are included.
4. Testing approach.
5. Realistic delivery timeline.
6. Total fixed-price quotation.
7. All exclusions and recurring costs.
8. Confirmation of source-code ownership and handover.
9. Post-delivery bug-fixing period.

I need a complete, credible, functional website that helps Latent Lens win real customers and establishes a strong foundation for an international AI technology business.

Please submit a detailed proposal, relevant portfolio examples, timeline and transparent quotation.