Construction — project intelligence & business development engine
Multiple active sites, a lab team handling specs, site managers chasing progress, procurement moving orders through a chain that never quite lined up. Everyone working hard. Nobody working from the same picture.
Managers kept project context in WhatsApp threads, email chains, and site-visit notes. The lab team got updates late or not at all — a concrete-mix question needing a same-day answer would sit until someone forwarded a screenshot. When a supplier delay hit, the site found out after crews were already mobilized.
Project knowledge existed — drawings, BOQs, approval history, vendor quotes, site logs — but it was scattered. A junior engineer asking "what did we finalize for façade spec on Block C?" meant paging three people. Senior managers had no lightweight way to check alignment without calling a meeting.
On the business development side, the team was drowning in opportunity noise. Tender notices arrived from portals, contacts, and forwarded PDFs. Qualifying a lead meant manually opening documents and guessing fit. Good opportunities were missed because they arrived late; weak ones ate prep time anyway.
The slowest part of bid prep was always the same: getting from 2D tender drawings to a usable quantity position. Someone had to sit with architectural and structural sheets, run manual takeoffs, and build a draft Bill of Quantities before pricing could even start. Off-the-shelf tools existed, but none matched how this company structured BOQ sections, described line items, or handed work from BD into estimating.
We approached this as three connected systems — live project coordination, upstream business development, and a custom drawing-intake layer that fed the BD desk.
AI workflows sat across the conversation and data layers, not inside a dashboard nobody would check on site. When a manager flagged a spec question or approval blocker, the workflow pulled relevant context and routed it to the lab queue with the right background attached — not a bare "please advise." Material-chain triggers linked BOQ line items to procurement status and site readiness, so a slipped cement delivery on a job with a pour in 48 hours escalated to the site lead and procurement instead of waiting for a weekly call.
A tiered-access knowledge base let anyone query by role and depth — a supervisor pulling today's approved drawing revision, a PM pulling full change history — with answers tied to indexed source records, not chatbot fluff. It deployed on WhatsApp, because that was already where managers, engineers, and coordinators talked. No new login, no training deck.
The engine continuously ingested market signals from public and licensed sources, normalized them into one opportunity feed, and scored each lead for fit against active sectors, geography, project size, and delivery capacity. Duplicates were collapsed and low-fit noise filtered before it hit anyone's desk. Each qualified opportunity arrived as a structured brief — scope summary, key dates, contract band, compliance flags, relevance score — so BD managers worked a daily prioritized queue instead of a raw dump. The team still made every bid/no-bid call.
A custom drawing-intelligence tool, tuned to this company's measurement conventions, read 2D tender sheets, identified measurable elements by trade section, and ran automated takeoff against their BOQ structure — substructure, superstructure, finishes, MEP allowances, external works — with standard units and descriptions aligned to how their bills are normally issued. Output was a draft BOQ ready for QS review, a takeoff abstract, and a scope summary. When a qualified tender landed with drawings attached, the workflow kicked off automatically and produced a full tender-intake bundle, so estimators started from a reviewed draft and spent their time on rates, risk, and exclusions — the work that actually wins or loses a bid.
| Metric | Before | After |
|---|---|---|
| Median manager-to-lab response (urgent) | 31 hrs | 6 hrs |
| Material delays caught before site impact | ~22% | 74% |
| Time to retrieve a verified project decision | 45+ min | <2 min |
| Weekly alignment meetings per project | 4.1 | 1.6 |
| Tenders reviewed per BD manager / week | 8–12 | 35+ |
| Bid prep lead time (ID to draft) | 11 days | 6 days |
| Drawing-to-draft-BOQ turnaround | 4–5 days | same day |
| QS hours on first-pass takeoff / tender | 28–36 hrs | 9–12 hrs |
| Tender win rate (qualified pipeline) | 14% | 21% |
| Duplicate / conflicting lookups / month | 19 | 4 |
Site teams stopped treating the lab as a black box. BD went from reactive inbox-clearing to working a ranked pipeline — fewer random leads, more time on winnable ones. The drawing tool didn't replace the quantity surveyor; it removed the blank-page problem. Bid meetings started with quantities on the table instead of a promise to "have the BOQ by Thursday."
The WhatsApp bot answered too confidently, too early. The first version pulled from partial indexes and gave clean, wrong answers on drawing revisions that hadn't been uploaded yet. Two engineers acted on outdated spec guidance before we added source citations and an "unverified — escalate" fallback. Adoption dipped for three weeks.
Manager-to-lab sync assumed consistent input. Some managers sent tight, labeled updates; others dumped voice notes and photos with no project tag. The lab started ignoring automated tickets that still needed cleanup. Three lightweight required fields in WhatsApp — not a full form — dropped the noise fast.
Material-chain triggers fired on incomplete BOQs. Automations ran before line items were validated, sending false escalations to procurement for materials not yet on order. We gated triggers behind BOQ sign-off and rebuilt confidence notification by notification.
The BD engine surfaced volume before precision. Loose initial scoring produced 40+ "qualified" leads a week, and the team ignored the feed entirely. Tighter fit rules plus a human review gate for borderline scores brought it down to an actionable daily queue. Quality had to come before coverage.
The drawing tool struggled with bad tender PDFs. Scanned sheets, missing title blocks, and mixed scales produced confident-but-wrong quantities — one draft double-counted façade area because a reflected ceiling plan was read as a floor layout. Mandatory drawing-register validation and a low-confidence flag made QS review a hard gate, not a nice-to-have.
Auto-generated BOQs used the wrong measurement convention once. The company priced some contracts elementally and others by trade; the tool defaulted to trade sections, and one bundle had to be restructured by hand. Mapping contract type at intake — not after generation — fixed it.
- Meet teams where they already work. WhatsApp wasn't the "enterprise" choice on paper — it was the only channel that would actually get used on site. Convenience beat architecture every time.
- AI handoffs need structure, not just intelligence. A small amount of required context at the point of capture saves a lot of cleanup downstream.
- Automate detection, not decisions, on the material chain. Flagging risk early worked; auto-resolving procurement on unverified data created alert fatigue.
- BD automation wins on qualification, not collection. The value wasn't "find more tenders" — it was to stop wasting senior time on tenders that were never a fit.
- Project knowledge systems live or die on source trust. Citations and escalation paths aren't polish on a live site — they're safety.
- Draft BOQ is not issued BOQ. The win was speed to a reviewable draft, not eliminating the QS. Teams that treated the output as final pricing got burned; teams that treated it as a head start saved days.
- Customize to how the company bills, not how the software industry bills. Building around their BOQ sections, units, and BD handoff is what made the tool stick.