Joist AI
A private AI platform that turns years of AEC proposal data into an intelligent knowledge system, so teams can draft, analyze and quality-check RFP responses 30 to 40 percent faster
www.joist.aiProduct Overview
What It Is
Joist AI is a private AI platform trained exclusively on a firm's own data. It reads years of past proposals and turns them into a searchable, conversational knowledge base organized around the firm's people, projects and clients, then helps teams draft, customize and quality-check RFP responses against it.
The Problem It Solves
Current Approach
AEC marketing teams face the same cycle with every RFP deadline. They spend hours hunting through shared drives, multiple CRMs and past proposals to find relevant content, essentially recreating the wheel for each response even when similar work has been done before. Information stays siloed across offices, departments and project management systems, while tight deadlines of 30 to 40 hours per proposal compound with multiple concurrent submissions. Quality risks multiply as content bounces between marketing, field teams and subject matter experts. The result is that teams have no time for strategy, spending their energy assembling content rather than crafting winning narratives.
How It Changes
Joist functions as a private ChatGPT trained exclusively on the firm's data. It reads through three to five years of proposals, processing more than 3,000 documents in under three hours, then creates a searchable, conversational knowledge base organized around the firm's people, projects and companies. Instead of spending 70 percent of the time hunting for content, teams get it instantly, then use Joist to draft, customize and quality-check responses while focusing on strategy and creativity.
Area of Focus
Joist serves architecture, engineering and construction firms across all sectors, focusing specifically on RFP and RFQ response and proposal development. Secondary applications include marketing content creation, knowledge management and business development enablement. The platform is particularly strong for public sector work at city, state, county and federal level, where requirements are complex and detail-heavy, and is also effective for private sector letter proposals.
Key Features
Intelligent Knowledge Base
Pain: When did I last write this? What was that project called? Who updated the file last?
Drag and drop three to five years of proposals, resumes, case studies, award submissions and boilerplate into the platform. The system annotates and classifies every page into distinct objects such as cover letters, technical writeups, project descriptions and bios, creating a search experience that drills down to the exact page, paragraph and word. When Joist encounters a person's bio across multiple proposals it groups them with their master resume, so the AI handles the organization rather than a person doing the heavy lifting. The chat interface takes natural questions such as when the firm last worked on a hospital project in the Midwest under $50M, and returns answers with their source documents. Content can also be transformed in place. Highlight old copy and give a simple instruction, such as making it 800 words or rewriting it from first person to second. The system pulls lookalike projects with AI-generated reasoning explaining their relevance, and generates lists of past work filtered by region, cost range, market sector or technical capability.
RFP Analysis and Smart Drafting
Pain: Massive RFPs with tight deadlines and complex requirements that need coordination across multiple teams.
Upload an incoming RFP of any size and Joist breaks it down within minutes, generating a summary, scope, key dates, evaluation criteria, the legal forms needed and a technical proposal outline. It automatically generates RFI questions by flagging conflicting information or unclear requirements, and identifies pain points through thematic analysis of what the client cares about most. It builds a compliance matrix to track requirements for final review, and supports custom templates so a firm can upload its own go or no-go criteria, risk assessment frameworks or legal review checklists. The smart draft workflow starts from the RFP plus optional go-by proposals, instructions and addendums. The system builds a complete storyboard with over 100 micro-prompts, one for each section, so the outline can be reviewed and adjusted before generation and strategic control stays with the team. Within minutes it produces a 40 to 50 percent rough draft by pulling from historical content and matching it to the RFP requirements, and recommends relevant past projects with detailed reasoning for why they fit. Multiple team members across offices can work on the same draft at once, with real-time visibility into who is working on which section. A central dashboard tracks project title, client, market sector, submission date and strategy notes. Custom personas allow different voices for public versus private work, or by market sector.
Quality Assurance and Compliance
Pain: Errors slip through when content bounces between marketing, field teams and designers under deadline pressure.
Compliance checking matches the draft proposal against the original RFP to pre-grade it before submission. The system flags missing requirements such as legal forms, mailing addresses or unanswered questions, and rates each section with constructive feedback, for example scoring risk management and naming what is missing. It is a final safety net against submitting a proposal that could be disqualified over an overlooked requirement. Beyond compliance, Joist detects spelling and grammatical errors, assesses win strategy, and identifies the firm's differentiators throughout the proposal. It provides coherence ratings by section so the narrative flows logically, and offers cross-proposal benchmarking comparing how the firm addressed a specific topic such as risk management or safety protocols across the last six months, surfacing stronger examples from its own history. Brand safety keeps responses within brand guidelines and writing style, based on the firm's historical content.
Used By
Target Customer Types
PCL Construction
$8.1B revenue, 12,000 employees
“Joist AI has become an indispensable tool for us. The AI-powered search is not just efficient, it's accurate and reliable. I can't emphasize enough how much time it has saved us.”
—
Woodard & Curran
$289M revenue, 1,000 employees
“Using Joist AI has provided our team with a unique professional development opportunity. Its flexibility to adapt to our needs has been gratifying, and suggestions for improvements are always heard and implemented, making it a rewarding experience for our team.”
—
Operating Locations
Primary market: the United States. Emerging markets in early testing: Canada, Ireland and the United Kingdom.
Implementation & Onboarding
Workflows impacted: Joist directly transforms RFP responses, proposal assembly, quality assurance and compliance checking, knowledge management, and marketing content creation. Secondary workflows that benefit include business development research, new hire onboarding, cross-office consistency, client preparation, and strategic insights drawn from historical performance. Adoption timeline: Day one. Upload past proposals. The system indexes in hours and search and chat are live immediately. Weeks one to four. Used on the first RFPs, brand voice tuned, boilerplate added, CRM setup begins. Months two to three. Coordinators and business development onboard, with collaboration and custom personas established. Months three to six. Rollout to project managers, architects and engineers, expanding into business development research and client prep. Six months and beyond. Templates refined, advanced analytics adopted, and the platform becomes the firm's knowledge hub.
Security & Privacy
Integrations
ROI & Value
Document review efficiency
30 to 40 percent time savings per proposal. An industry standard 30 to 40 hours reduced to 20 to 28 hours per RFP response.
Processing speed
3,000 proposals processed in under three hours, giving same-day knowledge base setup from years of historical content.
Capacity increase
30 to 50 percent more proposals pursued without adding headcount, so marketing teams handle more opportunities with existing resources.
Peace of mind
Reduced deadline stress, with fewer late-night emergencies and less burnout.
Standardization
Consistent quality across offices, through a unified knowledge base and shared best practices.
Knowledge retention
Faster onboarding, weeks rather than months, preserving organizational memory in a searchable database.
Founding Team
Rohan Jawali
Founder and CEO
Under Rohan's leadership, Joist AI has grown rapidly, earned industry recognition and secured compliance milestones while advancing AI adoption in the built environment. Before founding Joist AI he built deep industry experience in construction operations, innovation and virtual design roles, including a long tenure with Hensel Phelps where he led innovation and VDC initiatives.
LinkedInFunding
Total Raised
$4M
Stage
Seed
Investors