Free · Take home
The three AI prompts
Vendor-neutral. Paste into any AI tool. Each catches a category of exception Mike makes today: not orderable as written, applies or strike, single or pair.
From DHI All Access 2026 · Nashville
Take the tools. Use them Monday.
Three vendor-neutral AI prompts and the DHI Data Standards Proposal from Manoj's talk. Free to take home.
Session: Friday, May 29, 2026 · 11:15 AM, Nashville
The Tools
Three things to take home.
The prompts and the proposal are free. The Charter Partner program is for shops ready to deploy shop-aware for their team.
Submitted to DHI · Read it
The DHI Data Standards Proposal
An Executive Brief submitted to DHI leadership proposing an industry-convened catalog-layer standard for Division 08. Phase 1: a 3-month feasibility study scoped to door closers. ACES/PIES, adapted for door hardware.
Why DHI, not BHMA
Phase 1 scope & deliverables
Working schema sketch (closers)
Anticipated questions & FAQ
Limited · 2–3 partners
The Charter Partner program
Full shop-aware deployment in under 30 days. Catalog ingestion included. Branded for your team and the architects, AHCs, and customers you serve. Money-back guaranteed.
AI Assistant deployed for your team
Full custom catalog ingestion — waived
Branded version of the assistant
Roadmap input — first cohort only
Commit by Wed, June 3, 2026.
What Shop-Aware Looks Like
The catches your team makes today, built into the system.
A shop-aware assistant checks each schedule line against the manufacturer catalog, the door schedule, and your shop's history. Returns the 5% that need senior eyes. Skips the 95% that don't.
Product Advisor — Live
Q
“My customer wants to know the swing force of this closer when installed on a 250-pound door.”
A
For a 250 lb door using the 4040XP, you'll need to set the spring size to a 4 or 5 depending on door width and frequency of use. Swing force at this configuration falls in the 8.5–10 lbf range with the standard arm.
4040XP_Catalogue.pdf, p.5
More catches, every bid
Four kinds of exceptions that show up on every hardware schedule. Framed as the questions the people on your team ask themselves before the bid goes out.
The estimator
“Did we catch every line that's not orderable as written?”
Flags lines where 'LCN 4040XP' or 'VD 99 × Width' need 3–5 more values before they can ship.
The order writer
“Did we translate every AHC notation to an orderable part number?”
Returns full SKUs with handing, function, finish, and width filled in — or names what's missing.
The junior estimator
“Did we exclude every spec'd item that doesn't apply to the door type?”
Flags weatherstrip on interior office sets, smoke gaskets on non-rated openings, the catches Mike makes.
The AHC liaison
“Did we catch every door type and size?”
Cross-references door dimensions against hardware counts — flags 6'-0" singles that should be pairs.
Shop-aware, not generic
Trained on YOUR shop's data — your last 500 quotes, your house standards, your customer's keying program. The line between Google and Mike: shop-aware AI knows your business.
Rules in code, not in the LLM
Deterministic catches — finish-to-series, fire rating, code compliance — live in code where they never drift. AI handles ambiguity and flags it instead of inventing an answer.
Exceptions, not answers
The system surfaces the 5% of lines that need senior eyes. Mike reviews what only Mike can decide. Juniors handle the rest with the system at their back.
The Talk
From Hardware Schedule to Orderable MPNs.
A practical playbook for Division 08 hardware.
Manoj Tiwari, Co-Founder and CTO of Conversant AI, on the four no-standards that turn every bid into a translation project — and what shop-aware AI looks like on Monday morning.
DateFriday, May 29, 2026
Time11:15 a.m. – 12:00 p.m. local
WhereDHI All Access 2026 · Nashville, TN
Read the slides, share with your team, or reach out to Manoj directly.
Who We Are
Built alongside the people who know this industry best.
Our team combines deep AI expertise with real-world door hardware experience.
Kyle Harrison
Co-Founder & CEO
Product leadership at Google building products that serve billions of users. Expertise in scaling complex technical products from conception to global deployment.
Manoj Tiwari
Co-Founder & CTO
Google Research leadership advancing AI systems. Multiple patents. Published research in machine learning and deep expertise in building and scaling LLM-focused teams and productization.
Andrew March
Head of GTM
Decades of experience in technology and operational roles spanning B2B distribution. Has worked directly with leading distributors and online retailers of architectural and door hardware, and advised on technology and AI strategy across the industry.
“We've heard the pain of turning product intelligence into sales — repeatedly, from people across this industry. We're building the tools to fix it.”
Read more about usFAQ
Questions you might have.
Yes. They work with any AI tool — ChatGPT, Claude, Gemini, or whatever your team already uses. Built to be model-agnostic.
Vendor-neutral, not shop-neutral. They produce useful catches against any manufacturer's How-To-Order catalog, but they don't know YOUR shop's house standards, prior approvals, or customer keying programs. That's the shop-aware layer — and it's what a Charter Partner deployment delivers.
Generic AI knows a little about everything. Shop-aware AI knows YOUR door hardware business — your manufacturer relationships, your distributor's stocking patterns, your senior estimator's prior approvals, your customer's house keying program.
The free prompts above operate on public catalog rules. A shop-aware system operates on your last 500 quotes too. That's the line between Google and Mike.
A formal Executive Brief submitted to DHI leadership, proposing an industry-convened catalog-layer standard for Division 08 — the shared data plumbing that connects spec to schedule to PO to fulfillment across competing manufacturers. Modeled on ACES/PIES in the automotive aftermarket.
Phase 1 is a 3-month feasibility study scoped to one product family: door closers. Phases 2 and beyond are conditional on what the study finds. Conversant's preferred role is technical advisor at no cost, with the study itself conducted by a neutral third party DHI selects. The full document is linked above.
Full shop-aware deployment in under 30 days. We ingest your full product line, spec sheets, install guides, and technical bulletins — and we work with you to capture the tribal knowledge in your team's heads before it walks out the door.
Every estimator, order writer, and AHC liaison gets instant access. Web and mobile. Annual commitment. Money-back guaranteed against a 200-search threshold in the first 60 days — if your team uses it that much and you're still not satisfied, full refund.
Your proprietary data stays yours. We don't share it, sell it, or use it outside the engagement. Public catalog data — what's already on a manufacturer's website or in their published price books — is treated as public.