pages per submission
200
Dense tender dossiers, specs, annexes, and product constraints.
Traceable tender intelligence for lighting teams
Turn 200-page submissions into governed, defensible lighting matches.
HEIDITENDER reads tender dossiers, extracts structured requirements, applies governed rules, filters against real supplier catalog data, and ranks only products that actually exist.
Not a generic AI chatbot. A governed workflow for bid teams that need both speed and traceable justification.
manual search per project
3h
Spent crossing PDFs, spreadsheets, catalogues, and supplier portals.
day-one return
13x
On the first 100 submissions, the client cost is only a fraction of the saved engineering time.
Act 1 - The Problem / Act 2 - The Status Quo is Broken
Lighting engineers still stitch together bid decisions across documents, catalogues, portals, and memory.
A lighting engineer can spend three hours in front of a 200-page submission manually searching for matching luminaires. Repeat that across 50 projects a year and the hidden cost becomes 150 hours of expert time lost to manual search, endless cross-checking, and decisions that are hard to defend later.
Word filesExcel sheetsPDF cataloguesSupplier portalsHuman memoryEndless cross-checking
200-page
typical lighting submission
Specifications, annexes, norms, product schedules, and scattered supplier references.
50
projects per year
A realistic annual bid load for a firm running repeated manual search and checking cycles.
150h
expert time lost
Three hours per project becomes a recurring hidden cost at every tendering team.
What breaks today
Product data is fragmented, formats are inconsistent, and critical requirements often live in people's heads instead of a governed matching workflow.
Act 3 - The Innovation / Act 4 - The Technical Solution
Not a generic AI chatbot. A governed matching workflow for real lighting tenders.
The agent reads the submission, extracts requirements such as room types, lux levels, IP ratings, and norms, applies governed rules, and ranks products against the client's private rational database. The result is faster, more explainable, and more defensible bid selection.
Read the submission
The agent parses the tender dossier, annexes, and schedules into a structured starting point instead of leaving critical details trapped in long documents.
Extract structured requirements
Room types, lux levels, IP ratings, norms, controls, and other constraints are pulled into a machine-readable brief.
Apply governed rules
Explicit matching rules stay versioned, reviewable, and explainable so the workflow remains defensible for bid teams.
Match real catalog products
The system filters against the client's private rational database and ranks only products that actually exist in supplier data.
Why this matters
Every shortlisted product stays attached to the extracted requirement, the applied rule, the catalog evidence, and the final ranking logic.
Evidence Chain
Every recommendation arrives with a traceable justification.
Instead of a black-box answer, HEIDITENDER preserves the chain between what the tender asked for, which rule was applied, what the private catalog data actually contains, and why a product was ranked higher or lower.
Structured tender brief
- Room type: classroom
- Target illuminance: 500 lux
- IP rating: IP54
- Norm: EN 12464-1
Governed matching logic
- Hard: direct_ugr <= 19
- Hard: ingress protection
- Soft: efficacy and output
- Soft: control compatibility
Private database reality
- Normalized supplier attributes
- Only existing SKUs are considered
- Client rational database stays private
- Every filter remains inspectable
Defensible shortlist
- Passes hard constraints
- Soft score: 0.91
- DALI match retained
- Why ranked: glare + IP + control fit
Bid teams can explain why a product was kept, filtered out, or ranked lower without relying on memory or manual post-hoc reconstruction.
Act 5 - The Business Model
Immediate client ROI, with market economics that scale.
The client case is immediate: less expert time burned on manual product search and a better audit trail for bid decisions. At the same time, the market and revenue model are large enough to justify building a dedicated operating system for lighting tenders.
Client value
CHF 34'000
engineer time saved
Across 100 submissions.
CHF 2'500
client cost
Only 7% of the value delivered.
13x
return on day one
Immediate ROI once submissions start flowing.
Market and revenue
1'500
firms in Switzerland
Submitting lighting tenders annually.
150'000
submissions / year
Approximate market volume.
15'000
submissions at 10%
With roughly 150 clients in year one.
CHF 1'575'000
year-one total
Setup revenue plus usage revenue combined.
CHF 10'000 one-time database setup per client.
CHF 25 per submission processed.
From search to decision
Move from fragmented tender search to faster, more defensible bid decisions.
HEIDITENDER is designed for lighting teams that need speed, structure, and a justification trail they can stand behind when the bid decision matters.