> ## Documentation Index > Fetch the complete documentation index at: https://docs.trubuild.io/llms.txt > Use this file to discover all available pages before exploring further. # Technical evaluation > Score contractor technical submissions against weighted criteria, manage post-tender clarification rounds, and export committee-ready results. Technical evaluation is where owner engineers, discipline leads, and technical evaluators turn contractor proposals into a reviewed, weighted, defensible score. TruBuild accelerates criteria setup, proposal analysis, scoring drafts, and clarification impact review, while keeping the technical team responsible for the final judgment. The workflow is package-based. Each evaluation round is preserved, so the team can show exactly how scores moved after post-tender clarifications (PTCs). For how technical rounds sit alongside commercial rounds and the query register, see [Rounds and clarifications](/platform/rounds). If you do not see technical evaluation in your workspace, contact TruBuild support to request the feature for your organization. ## Who this is for | Role | Main responsibility in this workflow | | -------------------- | ---------------------------------------------------------------------------- | | Technical lead | Owns criteria, weightings, round readiness, and the final technical position | | Discipline evaluator | Reviews AI-drafted scores, adjusts them, and approves each criterion | | Procurement manager | Coordinates the timetable, the PTC cycle, and award evidence | | Owner reviewer | Reviews rankings, score movement, and scoring rationale | Editing an evaluation requires technical or full access on the package. Viewers can open the evaluation and read results, but scoring, uploads, and approvals are disabled. Disabled controls show the message "You don't have permission to edit this evaluation". ## What you need | Input | Format | Notes | | ---------------------------------- | --------------------------------------- | --------------------------------------------------------------- | | **Package with contractors** | In-app setup | The contractor list should match the tender return | | **RFP documents** | Document upload | Used to generate criteria and give the analysis scoring context | | **Criteria file (optional)** | Excel (`.xlsx`) or CSV | Use this if you already have a prepared rubric | | **Contractor technical proposals** | Document upload, one set per contractor | Native PDFs and Word files read better than scanned images | | **PTC documents and responses** | Document upload | Needed only when you run a clarification round | ## What you get * A weighted criteria structure, organized by scope with individual breakdowns * **AI-drafted scores** (0 to 100, to one decimal place) per contractor and criterion, each backed by evidence with document and page references * A **review workflow** where evaluators adjust scores and comments, then approve each one to lock it in * A **contractor leaderboard** ranked by total weighted score, with score deltas against the previous round * An **Excel results workbook** and a **Word report** for committee review ## Evaluation status at a glance Each evaluation moves through four statuses: | Status | Meaning | | ------------------- | ------------------------------------------------------------- | | **Setup** | You are still configuring criteria, documents, or contractors | | **Analyzing** | TruBuild is processing submissions and drafting scores | | **Ready** | Results are available for review and approval | | **Review Complete** | The evaluation is marked as finalized | While an evaluation is **Analyzing**, you see a progress view with the current stage and a progress track. If a run fails, TruBuild shows the error and a **Retry Analysis** button. Any previous results are preserved. ## Before TruBuild A typical technical evaluation on a mid-sized tender (5 bidders, \~25 criteria) takes 2 to 3 weeks of elapsed time and consumes 60 to 100 evaluator-hours. A senior engineer manually writes criteria. 1 to 2 days. Send the rubric to 3 or 4 evaluators in Word. Each evaluator reads each proposal cover to cover. Evaluators type scores and free-text justifications into Word docs. 2 to 3 days per evaluator. Someone copies scores from multiple Word docs into an aggregator sheet. 1 day. Evaluators meet when scores diverge. Sometimes scores shift, with no audit trail. Everything is recompiled into a Word or PowerPoint report. 1 day. **Pain points:** * Rubric inconsistency: every evaluator builds their own style * No way to verify each proposal was checked against every criterion * Aggregation errors from copy-paste transcription * No justification trail: "why did Vendor B get 70 on project management?" ## With TruBuild Generate criteria from the RFP, upload a criteria file, or start from the default template. Review, adjust weightings, and confirm. Drop each technical submission into its contractor's slot. TruBuild scores every contractor against every criterion and attaches evidence from the source documents. Each evaluator reviews the drafted score and evidence, adjusts where needed, and approves each criterion. Weighted totals roll up into the contractor leaderboard. Export the Excel workbook and Word report for the committee. ## Step-by-step: the initial submission round The first round of every technical evaluation is labelled **Initial Submission**. It sets the criteria and the baseline scores that later rounds compare against. ### 1. Upload RFP documents Upload the tender documents into the **Request For Proposals (RFP)** zone. TruBuild uses them to generate criteria and as scoring context during analysis. If your team already uploaded RFP documents in Tender Queries, the setup screen offers them here: "X RFP document(s) from Tender Queries". Click **Import** to copy them across instead of re-uploading. ### 2. Define evaluation criteria In the **Define Evaluation Criteria** step, choose one of three methods: TruBuild reads the uploaded RFP documents and extracts a criteria structure automatically. Progress moves through **Discovering files**, **Loading documents**, **Extracting criteria**, and **Finishing up**. Check the extracted scopes, breakdowns, and weights before continuing. Upload your prepared rubric as Excel (`.xlsx`) or CSV. TruBuild parses the file into scopes and breakdowns. Fix any misreads before continuing. If you start without documents, TruBuild offers a pre-built criteria set covering common categories such as Technical Capability, Resources & Team, and Project Management. Edit it freely to match your package. Criteria are organized on two levels: | Level | What it is | Example | | --------- | ----------------------------------------------------------------- | ---------------------------------------- | | Scope | A category grouping of related criteria | Technical Capability, Resources & Team | | Breakdown | An individual criterion with a title, a description, and a weight | Experience & Track Record, Key Personnel | The description is the scoring guidance the analysis follows, so make it specific. Weights support up to two decimal places and must total exactly 100% across all breakdowns. The UI validates the total and confirms "Total weight is valid" before you can continue. ### 3. Review the criteria The **Criteria review** step shows the full structure before analysis. This is the moment to freeze the evaluation basis: confirm scopes, breakdowns, descriptions, weights, and the contractor list. Mid-evaluation weighting changes are a leading source of post-award disputes, so get sign-off from the technical lead here. ### 4. Upload contractor proposals Upload each contractor's technical submission into their **Vendor files** slot. If a contractor submitted multiple volumes (technical, QA, execution plan), upload them all under that contractor. OCR scanned-only documents before uploading; native files produce better evidence. ### 5. Run the analysis Start the analysis. The evaluation moves to **Analyzing** and shows live progress by stage. For each contractor and criterion, TruBuild: 1. Reads the relevant sections of the contractor's submission 2. Compares the content against the criterion description 3. Drafts a score on the 0 to 100 scale 4. Attaches evidence with file references, including page numbers where available When the run completes, the evaluation moves to **Ready**. ### 6. Review and approve scores Results appear as a contractor-by-criterion matrix. For each cell, the evaluator sees the drafted score and its evidence. Evidence is marked as automatically extracted or manually added, and you can add your own. Adjust the score where your judgment differs, then approve it to lock it in. Treat the AI score as a draft: the evaluator's approval is the final word. When the team has finished, mark the evaluation as reviewed to move it to **Review Complete**. ### 7. Read the leaderboard The contractor leaderboard sits at the top of the results view. It ranks contractors by total weighted score and shows each contractor's position. Scope-level totals show where each contractor is strong or weak, which is often more useful to a committee than the single headline number. ## Running a PTC round After the initial round, raise post-tender clarifications for material gaps, then create a PTC round to assess the responses. PTC rounds are labelled **Round 1**, **Round 2**, and so on, and organization admins can standardize custom round labels in [Settings](/get-started/settings). Each round inherits the criteria from the previous round, so the scoring basis stays consistent. A PTC round has three setup steps: **Round setup**, **Response files**, and **Affected criteria**. Choose the PTC source. Select **Use previous evaluation results** to carry the clarification items TruBuild identified in the prior round, or **Upload issued PTCs** to import the clarification documents you issued to contractors into the **Issued PTCs** zone. Upload each contractor's clarification responses into their **Response files** slot. Add any additional clarification material under **Supporting files**. Map each PTC item to the criterion it affects, then decide the impact per contractor and criterion. ### Mapping PTC items to criteria When you upload issued PTCs, TruBuild parses each item, records the type of issue it raises (an information gap in the submission or an incomplete response), and maps it to a criterion with a confidence badge: **High confidence**, **Medium confidence**, or **Low confidence**. Check low-confidence mappings manually; the setup screen warns you with "X uploaded PTC item(s) were mapped with low confidence and should be checked". Rows that could not be mapped at all appear as unmatched, with the counter "X uploaded rows still need to be reviewed". You must assign a criterion to every unmatched row, and clear both warnings, before the analysis can start. ### Rerun or carry forward For each contractor and criterion, decide what the clarification response means for the score: | Decision | Meaning | Use when | | ----------------- | ---------------------------------------------------------- | ------------------------------------------------------------------------------------------------ | | **Rerun** | Re-evaluate the criterion in this round using the response | The response materially changes the score basis, for example a revised commissioning plan | | **Carry Forward** | Keep the previous round's score | The response confirms information without changing the evidence, for example a date confirmation | Each decision carries a rationale, so the round record shows why unaffected scores were preserved. TruBuild suggests decisions with a default rationale such as "The new response appears to change this criterion", and you can override any decision or edit the rationale text. Bulk actions speed up large matrices: **Select all for contractor**, **Select all for criterion**, or **Select all**. Avoid rerunning everything after every clarification. A response usually affects one or two criteria, such as programme or testing and commissioning. Selective rerun keeps the audit trail clean and the review workload small. ### Comparing rounds After a PTC round completes, each rerun criterion shows a delta badge next to its score: a green up-arrow for an improvement (for example "+2.5"), a red down-arrow for a decrease (for example "-1.2"), and a gray minus showing "0.0" for no change. The leaderboard shows total-score deltas too, so you can see at a glance which contractor gained or lost ground after clarifications. ## Exporting results Click **Export** in the results view and choose an output: A multi-sheet workbook. The **Export results** dialog asks you to "Choose which vendors to include in the workbook"; all contractors are included by default. A formatted technical evaluation report. It generates automatically after analysis completes and shows "(not ready)" until generation finishes. The Excel workbook contains: * **Overview sheet**: contractor rankings, total scores per round, and score deltas across rounds * **Per-contractor sheets**: criterion scores by round, evidence, comments, and per-criterion deltas * **Summary data**: scope totals, weights, and cross-round comparison The Word report includes an executive summary, the criteria overview, contractor rankings, detailed scores with evidence, charts (a donut of scores by scope and a score distribution), and recommendations where the analysis generated them. Pair these exports with the rerun and carry-forward rationale when preparing the award record. See [Award readiness](/platform/award-readiness) for what the committee pack should contain. ## Best practices Mid-evaluation weighting changes are the top source of post-award disputes. Use the **Criteria review** step as a formal sign-off gate. "Technical approach" is too vague to score consistently. Give each breakdown a description with concrete anchors, for example "100 = fully addresses all specified conditions; 50 = addresses most; 0 = does not address". For criteria that drive the award, read the evidence behind the drafted score before approving. The AI has no commercial context; your evaluators' adjustments are where the real value is. A mismapped clarification silently reruns the wrong criterion. Clear the low-confidence and unmatched warnings before starting a PTC round analysis. The carry-forward rationale is what makes selective re-scoring defensible. A committee reviewer should be able to see why every unchanged score stayed unchanged. ## Related How technical rounds sit alongside commercial rounds and the query register. Manage the clarification register that feeds your PTC rounds. The controlled evaluation model behind approvals and round history. Check whether the latest reviewed technical round is committee-ready.