Written from 5 named sources · Oct 3 · first result Unlocking Creative Intelligence: AI Research and Ideation Playbook for Roslu Design LLC Prepared for: Jennyfer Luna, Principal, Roslu Design LLC, Palisades Park, NJ Date: October 3, 2026 Executive Summary Roslu Design LLC is a young New Jersey studio (filed May 2025, active) that takes an event from concept through fabrication and installation, with floral artistry as part of the offer. As the sole principal, you carry design, building, installing, selling, and market research. Your limiting resource is hours, not ideas. The opportunity. PigKnuckle.ai is built to turn a plain-language request into a finished deliverable. That can be a research document, comparison table, risk summary, or set of next steps, with citations attached [1]. According to its published guide, it reframes a casual request into a structured brief, plans the steps, routes each step to a suitable model, checks claims against real sources, and assembles the result [1]. You do not need prompt-engineering skills. You do need to be specific about what you are trying to decide. Where it fits. Use it as a research and decision-support workspace. It does the gathering, comparing, structuring, and first-drafting around your creative work. You keep taste, pricing, safety, contracts, and client relationships. Three early-stage questions it can help with: Question PigKnuckle.ai deliverable Who should we sell to? Segment comparison and a recommendation of two priority segments How do we stand apart? Competitor map and three positioning options How do we turn ideas into booked work faster? Scored concepts, build and floral briefs, package and proposal drafts Three things to test, not assume: Whether your combination of concept, fabrication, installation, and floral is distinctive in NJ/NY is a hypothesis. The competitor research below is designed to test it. Time savings are hypotheses for the pilot to measure. They are not guarantees. Your state record lists the business category as "Other Public Sector," which does not describe event design. Confirm with the NJ Division of Revenue whether that classification needs correcting. Two things to verify before you start. PigKnuckle.ai's guide lists a Free plan of 15 credits per month (Fast quality tier only), a Starter plan at $19 per month with 75 credits (Fast and Optimal tiers), and higher plans above that. The guide says to confirm current terms before relying on specific numbers [1]. Its privacy summary says it does not sell personal information and routes prompts only to model providers that do not train on submitted content. The guide also says to confirm retention, deletion, and processor terms before uploading confidential material [1]. Treat those as claims to verify on the vendor's own pages. Until you have, use public information and de-identified descriptions only. How the guide works. It sets out four functional playbooks, then a 14-day pilot, then guardrails. Examples marked "illustrative" are format demonstrations only. They are not recommendations for what to build or whom to target. Your portfolio, past clients, pricing, capacity, and staffing were not supplied, so bracketed placeholders mark where your real details go. How to Work With PigKnuckle.ai These habits come from the platform's own guide [1] and shape every prompt below. Write briefs, not chat prompts. Describe goal, audience, constraints, format, and what you already know. The guide says "act as a..." preambles are unnecessary because the platform handles role and structure itself [1]. Review the reframed brief before approving. The platform shows what it understood before it spends credits. The reframing and planning stage is free, and this is your last free checkpoint [1]. One question per task. Split distinct questions into separate runs instead of one mega-prompt [1]. The playbooks below are staged for that reason. Use free revisions. The first three revisions per result are free. After that, each costs one credit [1]. Use them to ask for stronger sourcing or a corrected assumption. Ask for citations and gaps. Request sources, and end every brief with "list what you could not verify." The guide also warns that attached sources are not proof that the sources support the claim. Open them yourself [1]. Each client cycle feeds the next. Research shapes concepts, concepts drive build and sourcing briefs, those feed proposals, and lessons from installs improve the next research brief. Save every brief that works as a template. Add this standard block to the end of every brief: Geography: New Jersey and New York metro only. Label every statement as verified fact, reasonable inference, creative suggestion, or assumption. Prefer primary sources (company sites, venue sites, trade associations, government pages) and note publication dates. List sources you rejected and why. List what you could not verify. Do not invent client facts, testimonials, or results. The block matters because search noise is real. Research gathered for this guide returned a restaurant review page [2], an unrelated subscription-brand agency ranking [3], and a recipe page [5]. Explicit geography, inclusion criteria, and relevance checks keep that kind of material out of your results. PigKnuckle.ai Functional Playbooks Playbook A: Ideation and Thematic Concepting Goal: Move from a client type or opportunity to a scored shortlist of buildable concepts. Workflow Brief: client type, event type, venue style, season, guest count, budget band, signature techniques or must-use materials, venue restrictions. Generate: 8 to 10 concepts. Score: rank on distinctiveness, feasibility, cost control, asset reusability, photography value, and segment fit. Develop: take the top two or three into palette, build breakdown, and install sequence. Human veto: check against your real shop capacity, vendor lead times, and venue rules. Prompt framework 1: Concept generation Develop 10 event concepts for a [premium corporate product launch / luxury fall wedding / nonprofit gala] in the New York metro area, around [guest count], in a [venue type], in [month]. Budget band: [your range]. Use floral artistry, custom scenic fabrication, and immersive guest environments. For each concept give: creative premise, guest journey in sequence, major build elements, floral direction, lighting and material suggestions, production complexity (Low/Med/High), logistical risks, sustainability considerations, and commercial appeal. Venue restrictions: [for example no open flame, no ceiling rigging]. [Standard block] Prompt framework 2: Scoring Rank the 10 concepts on distinctiveness, feasibility, cost control, reusability of assets, photography value, and fit for [chosen segment]. Give a 1-to-5 scorecard and recommend the top 3 with reasons. Prompt framework 3: Portfolio spec concepts. This is valuable for a studio still building its portfolio. Generate six spec-event concepts Roslu Design could photograph for its portfolio. Prioritize concepts that demonstrate custom builds, floral design, installation capability, and strong social-media visuals without requiring a celebrity or major brand client. Mark production complexity and what a one-day photo shoot would require. Illustrative output shape (format only, not a recommendation to build): Concept Premise (illustrative) Hero build (illustrative) Floral direction (illustrative) Concept A Enclosed garden-like arrival Freestanding modular frames; size and materials set by your fabricator Light palette with textural greens; stems and availability confirmed with supplier Concept B Warm, sculptural minimalism Freestanding textured forms with niches; size to be determined Seasonal garden-style palette; stems to be confirmed Treat any stem, dimension, or build size in an AI draft as unverified until a supplier or fabricator confirms it. Playbook B: Custom Fabrication Specs Goal: Turn a selected concept into build logic you can hand to a fabricator. This is not engineering. Text-based AI is good at structure, sequencing, and checklists, and weak at precise geometry. Prompt for ranges, modularity, and questions. Work in layers: concept, then palette, then build, then install plan. Several short tasks produce more usable output than one giant prompt [1]. Prompt framework 4: Build options For the selected concept [paste concept summary], outline three custom build options: a statement entry, a focal installation, and a photo moment. For each give: candidate materials with two alternatives, approximate dimensions as a RANGE marked UNVERIFIED, modular breakdown for transport, load-in and install steps in order, estimated crew roles, failure points, and reuse potential. Venue constraints: [list]. Then list the questions a site visit must resolve: ceiling height, entry points, load-in route, power, rigging permissions, floor loading. Do not provide structural, rigging, or load calculations. Prompt framework 5: Budget tiers Provide Value, Signature, and Showstopper versions of the focal installation. For each, describe what changes in materials, florals, and labor. Do not estimate final prices. List the cost drivers for me to price myself. Space-first checklist. Before running Prompt 4, gather these so the output reflects reality: Venue type and ceiling height Entry points and sightlines Load-in route and window Power access Restrictions (flame, hanging, adhesives, scent) If any are unknown, say so and ask the tool to list the questions to resolve. Hard boundary: AI dimensions and descriptions are ideation only. Structural, rigging, and load decisions require a qualified fabricator or engineer. Playbook C: Floral and Material Sourcing Goal: Build a palette and a sourcing question list. The tool cannot see live wholesale inventory, so it cannot confirm availability, pricing, or order dates. Prompt framework 6: Palettes For the top-ranked concept, propose three floral and foliage palettes for [month] in the Northeast. For each, list hero stems, supporting greens, textural elements, and non-floral accents. Discuss typical seasonal availability, likely sourcing constraints, vase-life and perishability, allergen or scent flags, and how each palette reads in photographs under warm and cool lighting. Label all availability and vase-life statements UNVERIFIED hypotheses for me to confirm with NJ/NY suppliers. Do not provide order quantities or order-by dates. Prompt framework 7: Sourcing checklist Turn the chosen palette into a sourcing checklist: stems to confirm, acceptable substitutes if shorted, questions to ask the wholesaler, and storage and handling notes. This is a question list, not an order sheet. Prompt framework 8: Supplier and material discovery List wholesale floral suppliers, growers, rental houses, and fabrication-material vendors serving NJ and NY that publicly describe what they carry. For each give name, location, category, public contact path, and what they publicly say they offer. Separate observed facts from inference. List what you could not verify. You confirm final stems, quantities, pricing, and order dates directly with suppliers. No availability claim from an AI draft goes into a client proposal without supplier confirmation. Playbook D: Proposal Generation Goal: Shorten the path from discovery call to a polished, honest proposal. One-time foundations (run after you choose a segment and positioning, in Days 10 to 12 of the pilot): Prompt framework 9: Packages and capabilities Using this positioning for Roslu Design: [paste your one-sentence positioning], create three service-package structures for clients seeking event design, custom builds, and floral artistry. For each include the business problem it solves, scope boundaries, client inputs required, likely timeline, optional add-ons, and questions Roslu should answer before publishing prices. Also draft a one-page capabilities statement and a 15-question discovery-call questionnaire. Do not fabricate testimonials, past clients, portfolio results, or performance metrics. Prompt framework 10: Per-opportunity proposal Draft a proposal from these de-identified discovery notes: [paste notes]. Include: understanding of the client's goals, concept narrative (about 200 words), scope and exclusions, client inputs required, timeline with milestones, optional add-ons, and next steps. Distinguish client-stated facts from your suggested language. List missing information I must collect before pricing. Add answers to five likely objections on budget, timeline, and logistics, and describe a mood board in words so I can source licensed images. Prompt framework 11: Outreach and follow-up Draft five short outreach emails to [venues / planners / rental houses] introducing Roslu Design's [chosen positioning]. Each should reference only claims I can support: [list]. Add post-event case-study interview questions I can ask a client. The non-negotiable rule. PigKnuckle.ai's own go-to-market material argues that trust is central when selling AI-assisted work, and that every public artifact should carry verified claims or nothing [4]. Apply that standard to Roslu. Every statement about experience, capacity, insurance, staffing, or results must be checked against reality before it reaches a client. Low-Friction 14-Day Pilot Plan Design principles No integrations and no new software beyond PigKnuckle.ai. One folder for saved briefs and outputs. A clear continue, narrow, or stop decision on Day 14. Honest time budget: about 10 to 14 hours total over two weeks, in short blocks. That is heavier than a casual trial because the plan covers two research briefs, ten scored concepts, sales assets, and verification. If it is too much during event season, cut scope (one segment brief, five concepts, three outreach emails). Do not cut the verification steps. Credit budget: each run costs 1 credit on Fast, 2 on Optimal, 4 on Max, and planning is free [1]. The Free plan's 15 Fast-tier credits can cover a first trial. If Fast output looks too shallow, the $19 Starter plan adds the Optimal tier [1]. Confirm current plans first. Suggested calendar (Day 1 is Monday, October 5, 2026; shift freely): Days Phase What you do About time Done when 1 to 2 (Oct 5-6) Setup gate Create an account. Confirm current pricing, credit limits, file-upload options, export formats, and data-handling terms on PigKnuckle.ai's own pages. Run one small non-sensitive test task to see how the reframed brief and result look [1]. Record baseline hours for your last market scan, concept round, and proposal draft, if you have them. If you do not, time each pilot task instead. Define success. 1 to 1.5 hrs Gate passed or fallback in place; success criteria written 3 to 6 (Oct 7-10) Market foundation Run Brief 1 and Brief 2 below as separate tasks. Review each reframed brief before approving. Spot-check at least five claims against the original sources. 4 to 5 hrs One-page decision memo naming a primary and a secondary segment 7 to 9 (Oct 11-13) Positioning and concepts Run the positioning brief, choose one option, then run Prompts 1 and 2. Develop the top two concepts with Prompts 4 and 6. Check feasibility against shop capacity and vendor lead times. 4 to 5 hrs Two-concept shortlist and a draft positioning statement 10 to 12 (Oct 14-16) Sales assets Run Prompts 9 and 8 (vendor, venue, planner discovery) and draft five emails with Prompt 11. Delete any claim you cannot prove. 3 to 4 hrs Proposal template and an outreach list you have verified 13 to 14 (Oct 17-18) Test and decide Share the shortlist with two or three trusted planners, venues, or past collaborators. Send 5 to 10 outreach messages only if contacts and claims are verified. Feed feedback into one revision of the weakest output. Complete the scorecard. 1 to 2 hrs Decision: Continue, Narrow, or Stop Hard gate on data. Do not start Days 3 to 12 until you have reviewed the data-handling terms. If the vendor is slow to answer, use public-information-only prompts with no client names, contracts, personal data, or private pricing, or pause the pilot. Brief 1: Segment prioritization I'm a boutique studio in New Jersey offering concept development, custom fabrication, installation, and floral artistry for events. I need to decide where to focus for the next 12 months. Compare corporate events, weddings, nonprofit galas, retail activations, hospitality events, and public-sector events in NJ and NY. For each: typical buyer, project size, buying cycle, competitive intensity, seasonality, required capabilities, likely referral sources, and evidence of demand. Use named sources, give me a comparison table, distinguish facts from assumptions, and recommend two segments. [Standard block] Brief 2: Competitor map and positioning Create a competitor map for event designers, experiential-marketing studios, floral-production companies, and custom event fabricators serving NJ, NYC, and nearby markets. Include name, location, service categories, target clients, visible price signals, portfolio themes, differentiators, venues or partners mentioned, and evidence of recent work. Separate observed facts from interpretation. Test whether offering design, fabrication, installation, and floral together is actually uncommon, and identify underserved positioning. [Standard block] Brief 3: Positioning options Using the segment decision [paste memo], give three positioning options for Roslu Design. For each: target customer, core promise, proof points I need to develop, services to lead with and to avoid leading with, a website headline, and a sample outreach message. Do not assume portfolio or clients I have not listed: [list what you actually have]. Optional trend brief (quarterly). Ask for design motifs, materials, and seasonal floral trends relevant to NJ/NY events. Require every trend to be tied to a cited source and labeled observed in multiple sources, emerging, or speculative. Treat trends as inspiration, not as proof of demand. Making it stick. Save each working brief as a reusable template: a quarterly market scan, a per-lead proposal brief, and a pre-pitch concept round. The platform describes itself as a work-completion engine and not just a prompt environment [1]. The payoff comes from repeating the workflows. Pilot scorecard Leave baselines blank if you have none. That is normal for a studio that opened in 2025. Metric Baseline (if any) Pilot result Keep? Hours from brief to scored shortlist Hours from discovery notes to proposal draft Factual errors found in spot-check, out of claims checked Share of outreach targets correct and relevant after verification Outputs actually used in real client or partner contact Segment decision made (yes/no) Outreach responses (x of 5 to 10 sent) Credits used and plan cost Win rate and bookings cannot be judged inside 14 days. If you continue, review proposals sent against proposals booked over the following 60 to 90 days. Decision rule. Continue if the segment decision is made, errors are manageable, and at least one output was used in real contact. Narrow to one workflow if only one playbook paid off. Stop if outputs consistently need heavy correction. The pilot is designed to be reversible. Operational Guardrails and Best Practices A. Prompting for spatial and visual work Describe the space first. Give venue type, ceiling height, entry points, sightlines, load-in route, and power. Where unknown, say so and ask for the questions to resolve. Use ranges and flag them. Label dimensions "unverified; confirm with site visit and fabricator." Specify the guest journey in sequence: arrival, key moments, photo moments, exit. Name lighting and conditions. Warm versus cool light, day versus evening, and indoor versus outdoor change how florals and finishes read. Set constraints up front. Budget band, install window, venue restrictions, reuse goals. Ask for build logic, not just looks. Request modular breakdowns, transport and assembly steps, and failure points. Iterate in layers. Concept, palette, build, install plan. Use separate tasks, not one mega-prompt [1]. Use words as a mood board. Source real imagery separately, with copyright and attribution in mind. Treat renderings and drawings as ideation. A qualified fabricator or engineer decides structure, rigging, and load. B. Factual verification Four labels on every output: verified fact, reasonable inference, creative suggestion, assumption. Primary sources first, with publication dates. Relevance filtering: state geography and inclusion and exclusion criteria, and ask the tool to list rejected sources. The irrelevant results found while preparing this guide [2][3][5] show why. Open the sources. The platform's own guide warns that "sources attached" is not citation validation. Check that each cited page supports the claim, including dates and authorship [1]. Spot-check rule: open the original for any claim that drives a decision or appears in client-facing material. Never rely on AI alone for: competitor pricing, legal or insurance guidance, fire-code and safety requirements, venue policies, permit rules, or production cost estimates. No fabricated proof: no AI-drafted testimonials, past-client claims, or performance numbers [4]. Date-check trend and event information, because it goes stale quickly. C. Confidentiality and data Keep client names, contracts, personal data, and private pricing out of prompts until you have reviewed the vendor's data terms yourself [1]. Use de-identified discovery notes (for example, "a corporate client launching a product in fall") instead of names. Use the free reframing checkpoint to confirm the tool understood your request before spending credits [1]. D. Creative integrity and client trust Originality: use generated concepts as starting points and add Roslu's own design signature, so work stays distinctive and does not resemble others' designs by accident. Disclosure: be honest with clients about how AI-assisted work is used where it matters to them. Human control: you own all creative judgment, client relationships, contracts, safety decisions, and final commitments. Reversibility: if outputs routinely need heavy correction, narrow the use cases or stop. E. Quick reference: use the tool, or do it yourself Task Use PigKnuckle.ai Keep human or confirm independently Segment and competitor research Draft and structure Spot-check sources; judge fit Concept generation and scoring Generate and rank Taste, feasibility, capacity Build specs Structure and checklist Dimensions, rigging, load, safety Floral palettes Options and question lists Availability, pricing, order dates Proposals and outreach First drafts Every claim, price, and commitment Your First Three Moves Today to Monday: confirm PigKnuckle.ai's current plans, credits, and data-handling terms directly with the vendor, and decide whether the Free plan is enough to start. Before Day 1: write down your baseline hours if you have them, and gather what the placeholders need: a short list of your real portfolio pieces, your fabrication and installation capacity, and your target budget bands. Day 1 or 2: run one small non-sensitive task, review the reframed brief, and then start Brief 1. Roslu's advantage is a studio that can imagine, build, and install an environment. PigKnuckle.ai's job is to clear the research and drafting work that stands between your ideas and booked projects, while you keep every decision that carries your name. Sources [1] PigKnuckle.ai User Guide for Knowledge Researchers — https://pigknuckle.ai/share/pigknuckle-ai-user-guide-for-knowledge-researchers-Q7JTOavSZSKc [2] HAN SPOT - Updated October 2026 - 150 Reviews — https://www.yelp.com/biz/han-spot-gaithersburg-2 [3] Best DTC Creative Agencies for Subscription Brands 2026 – Apex Brands — https://apexbrands.io/blog/best-dtc-creative-agencies-for-subscription-brands [4] Pig Knuckle Go-To-Market Strategy — https://pigknuckle.ai/share/pig-knuckle-go-to-market-strategy-X8uYEanCQBE0