Craft the Brain of a Smart Style Assistant - LLM ChatBot (3-Week Sprint | Phase 1)

Craft the Brain of a Smart Style Assistant - LLM ChatBot (3-Week Sprint | Phase 1 with Vision for Phase 2)

Project Type & Budget:

Total Project Value: ₹1.5L

1. Phase 1 Budget: ₹50,000 (Fixed, All-Inclusive): 3-week sprint to build a working prototype with prompt logic, persona tone, API + integration-ready output
2. Phase 2 Budget: Up to ₹100,000: Scope and award contingent on successful Phase 1 delivery and user validation. Includes memory, wardrobe integration, and advanced personalization.

Timeline:

3 Weeks from project start, including integration support for iOS and Android apps (handled by internal team)

Project Overview:

We’re building a prototype conversational assistant using an LLM (e.g., GPT-3.5, Claude, DeepSeek, etc.) to simulate a friendly, on-demand stylist.

This Phase 1 sprint is focused on validating prompt logic, tone control, and real-world engagement - not building a full product (yet). The right output here sets the stage for a more advanced Agentic AI experience in Phase 2.

Note on Phase 1 Objective:

This is a Phase 1 prototype - not a full-stack build, but it’s mission-critical. The purpose is to test how users respond to a smart, conversational style assistant, and whether the interaction feels intuitive, engaging, and worth building on.

- If it doesn’t work well in Phase 1 - there’s no Phase 2.
- If it’s thoughtfully built and impactful - you’ll lead what comes next.

We’re not looking for shortcuts - we’re looking for a sharp foundation.

Core Deliverables:

1. LLM-Powered Chatbot Logic
• Modular prompt design for:
• Occasion, mood, vibe, budget
• Persona-based tone variation (e.g., 18–25, 26–35, 35+)
• Fallback flows (e.g., “Not this,” “Show me another”)

2. API Build + Output
• Build a clean, testable REST API
• Output in structured JSON:
• Outfit suggestion
• Justification for the choice
• Optional CTA (“Save this,” “More like this”)

3. Integration & Support
• Work with our mobile dev team to test integration
• Provide endpoint documentation and sample payloads
• 30-day post-launch mandatory support (bug fixes + prompt refinements only)

Optional Bonus – Hindi Support (1 Flow):

If you’re confident in building multilingual prompt logic, we’d love to see one sample flow in Hindi included in the response architecture. This is not mandatory, but will:
• Help us assess your multilingual flexibility
• Serve as input for deeper localization in Phase 2
• Give you an edge if we go multilingual at scale

Prompt Logic Approach (for Developer Context):

We follow a modular prompt design philosophy inspired by proven frameworks like:
• R–T–F (Role–Task–Format)
• T–A–G (Task–Action–Goal)
• C–A–R–E (Context–Action–Result–Example)
• R–I–S–E (Role–Input–Steps–Expectation)

You’re free to adapt the structure, but we expect clean, interpretable prompts that support variation, tone control, and future personalization.

Mandatory Conditions:

• ₹50,000 fixed budget
• 3-week sprint with weekly milestones
• GitHub-based development from Day 1
• All IP to be mandatorily assigned to us prior to payment
• No external hosting or shadow repositories

Project Monitoring:
• Week 1: Prompt architecture + tone mapping
• Week 2: Core build and testable API
• Week 3: Final integration support + refinements
→ Daily or alternate-day updates via Freelancer or GitHub expected

Additional Clauses:

1. Scope of Work Disclosure: Full SoW and flow logic shared only with shortlisted candidates. A basic test version demo will be required before final selection.
2. Confidentiality Agreement: Final selection requires signing an NDA and Consultant Contract (templates will be provided).
3. Intellectual Property Rights: All work delivered must be original and exclusive. We retain full rights to use, modify, and extend the chatbot logic in all future phases.

Important Note on Selection

This will be a sharp and structured selection process.

We’re only looking to partner with:
• Builders who’ve worked on LLM-driven bots
• Engineers who understand prompt design and persona-based tone
• Developers who build clean, testable, API-first logic
• Teams who can deliver on time, on GitHub, and within budget

No general-purpose bots. No vague portfolios.
Demo links and real-world examples are mandatory.

Phase 2 Opportunity:

If Phase 1 performs well, we will directly engage the selected partner for Phase 2 - a deeper build involving:
• Conversational memory
• Wardrobe integration
• Contextual nudges and multi-session continuity

Budget up to ₹100,000 (to be finalized based on adoption metrics and user feedback).

Why This Build Matters (And Why You Might Be the Right Fit):

This is more than a chatbot. It’s your chance to build a signature Agentic AI assistant that feels intuitive, emotionally resonant, and behaviorally sharp.

This is the kind of project where:
• The prompt logic carries your fingerprint
• The conversation flow reflects your thought process
• And yes - you may even get to name the chatbot as part of your creative ownership of Phase 1

Your code didn’t just work - it spoke. And it wore your signature like a badge of honor.

We want someone who:
• Thinks beyond code; and builds conversation-led experiences
• Wants to create something differentiated and lasting
• Sees Phase 1 as the start of a bigger, bolder journey

To Apply:

Please include:

1. 1–2 demo links of past chatbot or assistant work
2. Your preferred LLM stack (GPT-3.5, Claude, DeepSeek, etc.)
3. Confirmation of:
• GitHub-based development
• Fixed ₹50,000 budget (end-to-end, Phase 1)
• 3-week sprint
• NDA + contract readiness

Let’s build something thoughtful, creative, and sharp.
We’re excited to get started.