Contact usENعربي

Case studyDefence industry

SAMI: Designing Bid Intelligence for Government Tenders

Client

Tender pack

RFP-117p. 37 of 112

4. Special conditions

4.1

The contractor shall submit a monthly maintenance report within five working days of month end.

BoQ 14
4.2

The contractor shall keep three certified technicians on site for the full contract period.

No line for §4.2
4.3

Spare parts shall be supplied from the manufacturer's approved list.

BoQ 13

RFP · §4.2 · p. 37

Bill of quantities

  1. 12Preventive maintenance visits
  2. 13Spare parts
  3. 14Monthly reporting
  4. ?No line for §4.2

No matching BoQ line: flagged for pricing

Confirmed by the bid reviewer

Illustrative tender page with fictional text.

Saudi Arabian Military Industries (SAMI) competes for government tenders, many published on Etimad. Ting designed an AI Solution Blueprint for its bid workflow: one controlled workspace from the first look at a tender to the submitted bid.

Industry
Defence and aerospace
Location
Saudi Arabia
Scope
Etimad tender workflow, from intake to submission
Challenge
Long tender packs and many handoffs against a fixed deadline

By the Ting teamUpdated

Quote registerRFP-117
  1. 12Preventive maintenance visitsSupplier A · 02 Sep · Valid to 30 OctCurrent
  2. 13Spare partsSupplier B · 14 Jul · Valid to 14 AugExpired: re-quote
  3. §4.2Certified technicians on siteSupplier C · 09 Sep · Valid to 09 NovAdded after the BoQ check

Cost sheet

  1. Reviewedv4
  2. Approvedv4
  3. Submittedv4

Approved and submitted on the same version

Illustrative register with fictional suppliers and dates.

The challenge

Every government tender sets off the same chain. Someone finds the opportunity, several departments assess it, technical and pricing teams build the response, and approvals follow, all against a deadline that does not move. Many of these tenders are published on Etimad, the Kingdom's procurement platform.

A tender pack is long, and the details that matter most are not always where you expect them. In large tenders, obligations often sit in the special conditions rather than in the bill of quantities: a requirement for people, equipment or a certificate that no line item names, and which is therefore easy to leave unpriced.

Supplier quotations arrive on their own schedule, and a costing sheet can change after it was reviewed. With short tender windows, every repeated handoff eats into the time left to build a strong response.

Our approach

  1. One record per tender. Each opportunity is registered once, with its source, deadline, documents and accountable owner, instead of being rebuilt in every department's files.

  2. AI that shows its source. The workspace extracts dates, scope, eligibility, penalties and special conditions from the RFP. Every extracted point links to its document and page, and a reviewer confirms or corrects it before it affects a score, a cost or a word of the proposal.

  3. A bid decision with evidence. SAMI's own scorecard becomes the rule. Each function enters its input, and a named person records the decision, go, no-go or an exception, with the reasons.

  4. Costs that trace back to the tender. A requirement-to-BoQ matrix catches obligations that no line item names. A quote register keeps each supplier price with its date and validity, and a versioned cost sheet ties every approval to the exact version approved, so no one signs one number and submits another.

  5. A pilot that starts small. Walk one safely redacted real tender through the process, then pilot the bid cockpit on one repeatable tender category before adding cost, compliance, bid memory and integrations. Throughout, the assistant may not submit bids, contact suppliers, set prices or approve risk.

What was delivered, and how it will be judged

A Blueprint, handed over for SAMI's internal review. What a pilot has to prove is set out below; no pilot has run.

  1. What SAMI received30 September 2026

    The AI Solution Blueprint, handed over for SAMI's internal review: the target workflow from tender intake to submission, six capability areas, the data and control design, and a staged pilot plan.

  2. What a pilot has to prove

    The time to a verified bid decision, and the share of prices backed by a current supplier quote. Targets are set once SAMI confirms its own baseline, and the Blueprint makes no promise about win rates.

  3. Later, where policy allows

    Searchable bid memory from past tenders and permitted public award records, then approved connections to Etimad, supplier communication and internal systems.

No pilot has run yet, so no time-saving or accuracy figure is published.

Why this matters

Bidding is one of the few processes where a company's knowledge, discipline and speed are tested against a hard deadline every time. When the work lives in inboxes and spreadsheets, the cost is rarely visible: an obligation nobody priced, an approval given on an older version of the sheet, a strong response squeezed into the final hours.

The Blueprint does not describe an AI that writes bids. It describes a bid process that can be checked.

For any Saudi company that sells through tenders, that is the order that works. A faster process nobody can audit is not an advantage in a regulated market.

More case studies

All case studies

Sound like your operation?
Let's map yours.

Contact usSee the AI Solution Blueprint