From managing the bid to doing the bid

For most of Altura's existence, we built software to help bid managers manage the bid. An overview of everything being worked on, surveys to qualify each bid, evaluations based on customer feedback, and dashboards built on all that data. The promise behind it was compounding value: qualify and evaluate every bid in Altura, and over time, you'd build a database that gives you the insights to make smarter decisions for future bids.

In practice, it was hard to get right. The value only showed up after you'd fed the system many bids, and that feeding was manual work: finding the relevant information somewhere in your documents and inboxes, filling it into surveys and fields by hand, and keeping it all up to date as the bid moved on. For a bid manager under deadline pressure, that was extra admin from the very tool that was supposed to reduce admin. Meanwhile, the spreadsheets and emails never really went away.

But even for customers who did it all flawlessly, and while the manager’s got more strategic insights, the bid manager's week looked exactly the same. They were still reading 200-page tender documents to figure out what was being asked. Still pulling requirements out of walls of text, one by one. Still combing contracts for risky clauses. Still drafting clarification questions at 11 pm. And still spending the largest share of their energy chasing subject matter experts. Writing the proposal was never the real challenge in bid management. The real challenge is getting input from busy experts, each holding a piece of the answer, and none of them having time (or appetite) to dig through tender documents. Perfect project management didn't remove any of that. It just organised the pain.

So somewhere in 2025, we admitted to ourselves that our promise depended on humans feeding the system, and that part was never going to work. Now, if AI could do the work itself, the feeding would come free with it. It was time to stop managing the work and start doing it.

What "doing the work" actually means

The new Altura is built around workflows. Workflows do the work themselves: analysing documents, extracting and structuring information, checking flagged items against your own standards, drafting questions and answers, keeping everything up to date when new information arrives. The results are ready for review. There's no chatbot to prompt either. Each workflow is purpose-built for a task that every complex bid goes through anyway.

Here is what that looks like over the course of a bid. On day one, you upload the tender package and get a structured summary: key dates, scope, budget, evaluation method. The whole team starts from the same facts instead of their own mental models of 200 pages. Requirements extraction then pulls every requirement out of the documents into a clean, categorised list with source references.

From there, the compliance check matches each requirement against your company's documents and highlights any gaps before you commit. Risk analysis scans every clause against your organisation's risk framework and flags the ones your legal team would lose sleep over. When something is vague or unreasonable, Altura drafts the clarification questions. And before you submit, a final check runs your package against every obligation in the tender.

The chasing changes too. Instead of forwarding the tender package to your experts and hoping the right input comes back, each subject matter expert gets a short list of the items that need their judgment, each with a first answer plus the reasoning and sources behind it. They correct and confirm, rather than digging through documents to figure out what applies to them.

Each workflow feeds the next. The analysis feeds the risk assessment; risks turn into clarification questions, and the answers you receive flow back into your requirements and compliance picture. That used to be the bid manager's job too, by the way: keeping it all in sync manually every time a new document dropped.

That old promise of compounding value finally gets kept, too. Every answer and decision now land in the platform as a byproduct of doing the work, so the database that makes your future bids smarter builds itself. Nobody fills in a survey. Our Bid Companion (AI Agent) is available to help you make informed decisions by asking it questions; additionally, it can make edits for you on the platform.

Where the human went

The bid manager is still in the middle of everything. Every output is delivered as a draft with sources linked back to the original text, ready for them to review, adjust, and approve. You decide which data the platform uses. AI-generated content is marked as such. We built it this way because a wrong compliance answer in a tender is expensive, and because the bid manager's judgment is exactly the part worth protecting.

I started this company on the belief that bid management should be the strategic powerhouse of a commercial organisation. Bid managers have always had the talent for that. The job just left no room for it, because the reading, extracting, checking, and chasing consumed everything. When workflows take over that layer, what remains is the work that wins bids: deciding what to go for and why, shaping the win themes, and getting the right people to care long before the deadline forces them to.

That's the job most bid managers thought they were signing up for. After seven years, I believe we're finally building the product that lets them do it.