How we work
Every stage of the bid, mapped before anything is automated.
We start from the way your team really bids. Then we decide, stage by stage, where an agent helps, where a process change helps more, and where a person must stay in charge.
- Tested agent roleRoles we have run on tender work
- Configured in your processWe design this into your process
- Human-ownedPeople decide. Always.
The bid lifecycle
Twelve stages, six phases.
This is the process we map and instrument. It is not a claim that every stage is a finished agent product. Each stage is labelled honestly.
Phase
Find
- 01
Discovery and opportunity monitoring
Portals and notices watched against your service lines, regions and value bands.
- Tested agent role · Tender Scout
- Where agents help
- A weekly digest of new opportunities, with filter reasons for everything excluded.
- What people own
- Which criteria count as a fit, and whether to look closer.
- 01
Phase
Decide
- 02
Qualification and bid / no-bid
Fit, deliverability, capacity and mandatory risk weighed before anyone writes.
- Tested agent role · Bid Director
- Human-owned
- Where agents help
- A triage note: pursue, hold or do not pursue, with reasons and a suggested owner.
- What people own
- The go decision. Always.
- 02
Phase
Prepare
- 03
Portal, notice, deadline and document register
Frozen pack versions, clarification deadlines, addenda and portal rules in one register.
- Configured in your process
- Where agents help
- Flags when a pack changes or a clarification deadline is close.
- What people own
- Portal access, clarification questions and what version is final.
- 04
Tender breakdown
Requirements, lots, scoring and instructions pulled out of the pack.
- Tested agent role · Compliance
- Where agents help
- A structured list of every requirement, typed and referenced back to the page.
- What people own
- Interpretation where the buyer is ambiguous.
- 05
Compliance and evidence library
A matrix of requirements against approved evidence, with owners and expiry dates.
- Tested agent role · Compliance
- Configured in your process
- Where agents help
- Covered, partial and missing rows made visible before drafting starts.
- What people own
- What counts as approved evidence, and who maintains it.
- 06
SMEs and work allocation
Owners and reviewers for each section, checked against concurrent bids.
- Configured in your process
- Where agents help
- Routing suggestions and load warnings across live pursuits.
- What people own
- Who is asked, and what they are taken off.
- 03
Phase
Write
- 07
Planning and win themes
The story: why you, why now, what the buyer is really scoring.
- Human-owned
- Where agents help
- Pulls buyer context and past approved themes for the team to work from.
- What people own
- Strategy and win themes. Human-owned by design.
- 08
Drafting with source-grounded retrieval
Answers written from the approved library, with every claim traceable.
- Tested agent role · Bid Writer
- Where agents help
- Drafts only where a named approved source exists. Gaps are left open, not guessed.
- What people own
- Tone, judgement and anything the library does not cover.
- 09
Commercial and pricing
Rates, assumptions and commercial position.
- Human-owned
- Configured in your process
- Where agents help
- Checks the pricing schedule is complete against the instructions. It does not invent prices.
- What people own
- Every number. Pricing is human-owned.
- 07
Phase
Submit
- 10
Review cycles
Named reviewers, claim safety, and a red team on what is still open.
- Tested agent role · review handoff
- Human-owned
- Where agents help
- Review packs that show sources, open gaps and changed claims since last round.
- What people own
- Accept, edit or reject. Every time.
- 11
Submission pack
Final files checked against the instructions, approved and uploaded.
- Human-owned
- Configured in your process
- Where agents help
- A pre-submission checklist against format, word limits and required documents.
- What people own
- Approval and submission. Agents never log in to portals or submit.
- 10
Phase
Learn
- 12
Debrief, learning, renewals and pipeline metrics
Feedback captured, the library improved, renewals watched, capacity measured.
- Configured in your process
- Where agents help
- Debrief digests, renewals watchlists and capacity dashboards from your own data.
- What people own
- What changes as a result.
- 12
Engagement model
Workshop, configure, pilot, hand over.
A defined engagement with an end. Pilots start on approved sample or synthetic material and are measured against your own baseline.
- 01
Workflow discovery
We sit with your bid manager and the people who draft and review. We map how a tender really moves, where evidence is hunted, where review stalls, and what data may go where.
Process map, baseline notes, data and procurement constraints
- 02
Configure
We structure your approved knowledge library and configure Grok Bot agents in a workspace your organisation owns, with instructions and review gates written in plain language.
Library structure, agent instructions, review checklists
- 03
Pilot
Your team runs the workflow on approved sample or synthetic material first, then on a live pursuit once data owners say so. We tune instructions, not your prose.
Measured pilot against your own baseline
- 04
Train and hand over
The people who will run it practise briefing agents, rejecting weak drafts and filing accepted answers back. Then it is yours. Further support only if you want it, scoped separately.
A team that runs the next tender without us
Talk to us about your bid process
A discovery discussion with your bid manager or proposals lead. We ask how a tender moves through your team today, where approved knowledge lives and who signs things off, then tell you honestly whether in-house implementation fits.