Olra, the AI Platform for Public Construction Tenders

Responding to a public tender is time-consuming. You have to monitor the platforms, read 80-page procurement packs, write a technical proposal that precisely addresses the scoring criteria, verify that all administrative documents are in order, and assess your chances of winning before investing two full days into the bid. For a small contractor or construction SME, that workload is often simply too heavy. Olra is a French AI platform that covers the entire process, from tender detection through to proposal audit before submission.

Olra is built for tradespeople, micro-businesses, and SMEs in the construction and building trades that regularly bid on public contracts without a dedicated in-house team. The tool does not promise to bid on your behalf, but it does significantly reduce the time spent on each submission and improve the quality of the proposals you submit. Here is what the platform concretely does, who it is relevant for, and where its limits lie.

What Olra Is, Exactly

Olra is a French SaaS platform developed by Tensoria, purpose-built for responding to public procurement contracts in the construction sector. Its positioning is clear: cover the entire process, from daily tender monitoring through to proposal submission. It is not a generalist AI writing assistant like ChatGPT applied to tenders, nor a simple tender comparison tool. It is a suite of 6 interconnected modules designed around the real workflow of a tradesperson or construction SME.

The platform monitors 8 major French public tender sources (BOAMP, JOUE, PLACE, e-marchespublics, AWS-Achat, France Marchés, Klekoon) and covers contracts from €25,000 excl. VAT. Modules can be used independently or in sequence depending on which phase of a bid you are in. Hosting is 100% French (OVHCloud Roubaix), data is not used to train third-party models, and GDPR compliance is documented.

Olra is available at olra.fr with a 14-day free trial, no credit card required.

The 6 Modules in Detail

Olra's advantage over generalist alternatives is its end-to-end coverage of the bid cycle. Here is what each module concretely does and the real gain it provides for an organization without a dedicated team.

Tender monitoring

The first challenge for a contractor who wants to bid on public work is finding the right contracts. Official platforms (BOAMP, PLACE) are comprehensive but noisy: thousands of RFPs published daily, with titles that do not always reflect the actual scope of the work.

Olra addresses this through semantic matching: the algorithm does not simply filter by CPV code or keywords — it analyzes the full text of each notice to assess actual relevance against your business profile. The result: a daily email with 3 to 5 curated tenders, each with a relevance score. The platform also detects re-published tenders after an unsuccessful first round, which tend to be less competitive opportunities.

Monitoring is only available in the Pro plan (€149 excl. VAT/month). The Essential plan at €79 excl. VAT/month does not include this module. More information on monitoring features at olra.fr/veille-appels-offres.

Strategic tender-pack brief

Once a contract is identified, the first question is: is it worth bidding on, and if so, how should we position? Answering that requires reading the entire procurement pack (DCE), which often includes consultation regulations, technical and administrative clauses, and multiple annexes — sometimes 80+ pages in total.

The tender-pack brief module automatically extracts the key elements in under 2 minutes after upload:

• The weighted scoring criteria from the consultation regulations (technical quality, price, lead time, etc.)
• Implicit sub-criteria the buyer does not always state explicitly (occupied-site constraints, proximity to schools, sensitive zones)
• Flagging of risky contract clauses (excessive retentions, disproportionate penalties)
• A checklist of mandatory documents to provide
• A top-5 list of critical points to address in priority order

For a contractor evaluating 10 to 15 contracts a month, this module alone can save several days of work. The ability to detect unfavorable clauses and implicit buyer expectations is particularly valuable: it prevents you from bidding on poorly calibrated contracts or losing points on criteria that were never spelled out.

DECP competitive analysis

The DECP (Données Essentielles de la Commande Publique — Essential Public Procurement Data) is open data published by public buyers on their past contract awards. Olra uses this database to provide competitive context before you write your offer:

• History of comparable contracts over the past 24 months
• Observed median price (a reference for calibrating your bid)
• Top contract winners and their win frequency
• Identification of recurring buyers in your geographic area

This module answers a concrete question: what price point makes you competitive without undercutting the market? The data exists in public records, but collecting it manually takes hours. Olra does it in seconds.

AI-assisted technical proposal writing

The writing module is the most anticipated and the one that deserves the most honesty about what it actually does. Olra generates a structured first draft in 3 minutes, with 5 sections aligned to the consultation regulations, paragraph-by-paragraph traceability to the sources used, and a structure that addresses the weighted criteria identified in the brief.

This first draft is not a finished proposal. It is designed as a working base that the professional must enrich and adapt: their specific site references, their proprietary methodology, their actual staff and equipment. The AI generates structure and generic content; human review is essential for the differentiating elements. The writing module is available in early access for Pro and Enterprise subscribers at olra.fr/redaction-memoire-technique-ia.

AI audit of the technical proposal

This is probably Olra's most robust module. Once your proposal is drafted — whether with the tool or on your own — the audit analyzes the document on 14 critical points:

• Coverage of each sub-criterion in the consultation regulations
• Consistency between the schedule, staffing and equipment, and the references presented
• Wording issues likely to affect the score
• Identified strengths and gaps with explanations

The audit report is actionable: it does not just flag "insufficient on this criterion" — it explains why and suggests how to fix it. The claimed impact is an average improvement of +3 points on the technical quality score.

Weighted bid score simulation

The final module simulates the expected score your offer is likely to receive, taking into account both the technical quality (as assessed by the audit) and the price relative to DECP data. The simulation returns a score range (e.g., 12/20 to 17/20 on technical quality) and identifies 2 to 3 priority levers to improve the score before submission.

This module helps you make a concrete decision: submit as-is, improve on the identified points, or determine that the effort-to-success-probability ratio is not favorable.

Who Olra Is Right For

Olra is built for a specific profile. Here is an honest analysis of where the tool delivers real value and where it is not the right solution.

Profiles where Olra makes sense

A tradesperson or small construction company responding to 2 to 5 public contracts per month. This is the core target. For a plumbing, electrical, painting, or landscaping business that regularly bids on public work without a dedicated administrative assistant, Olra pays for itself on the first contract won or the first badly calibrated bid avoided.

A construction SME of 5 to 15 employees with a commercial manager handling tender responses. At that volume, the automatic tender-pack brief and the proposal audit represent several days of work saved per month. The score simulation also helps select which bids are worth investing time in.

A company that wants to improve its public tender win rate. If you already bid regularly but your success rate is low, the proposal audit can identify systematic gaps in your submissions (missed criteria, recurring inconsistencies) that you cannot see from the inside.

Cases where Olra is not the right solution

A contractor who bids occasionally on 1 or 2 contracts a year. Olra's monthly cost (€79 to €149 excl. VAT) is not justified for that volume. A manual approach using ChatGPT or Claude with your own business context will be sufficient.

A company that responds exclusively to private-sector tenders. Olra is centered on French public contracts. Private RFPs, architectural competitions, or off-public-procurement consultations are not covered.

A large company or enterprise with a dedicated tender team. Olra is sized for micro-businesses and SMEs. For high-volume operations with multiple response teams, a custom solution or an enterprise platform will be more appropriate.

Olra vs the Alternatives

The market for AI tools in public procurement took shape in 2025–2026. Here is how Olra compares to the main alternatives available.

ChatGPT, Claude, and Mistral used manually

Generalist AI tools can help write a technical proposal if you supply the right context. The problem is fragmentation: no connection to tender sources, no automatic extraction of criteria from the consultation regulations, no structured audit, no score simulation. These are building blocks you assemble by hand.

Olra is more constrained (dedicated interface, imposed workflow) but more complete and more reliable for regular use. For 1 or 2 tenders a year, generalist AI is enough. For 2 to 5 per month, the productivity difference is real.

Specgen

Specgen is a French tool that writes directly into your existing Word and Excel files, preserving your document formatting. It is very strong on drafting all contractual documents (technical specifications, proposals, pricing schedules) and is better suited to engineering consultancies and technical SMEs with professional formatting requirements.

The key difference from Olra is scope: Specgen is focused on writing; Olra also covers tender monitoring, competitive analysis, and post-drafting audit. If your primary need is to produce documents quickly in your own Word templates, Specgen is the better fit.

TenderNow

TenderNow offers a pipeline management approach with a reference library and proposal generation. The tool is more oriented toward managing multiple bids simultaneously than toward auditing and improving proposal quality.

Custom RAG solution

A different option is to build a custom internal AI assistant based on RAG, trained on your own past winning proposals, your company profile documents, your project references, and your resource catalog. This approach is more powerful for personalization (your writing style, references, and methodology are genuinely yours) but requires an initial configuration investment.

Olra and a custom RAG solution are complementary rather than competing: Olra handles monitoring and tender-pack analysis; a custom RAG excels at highly personalized drafting. For SMEs with a high bid volume and a strong history of winning submissions, combining both is an effective approach.

Limitations to Know

An honest assessment requires acknowledging what Olra does not (yet) do and the conditions for success.

Writing always requires serious expert review

The first draft Olra generates is a foundation, not a finished proposal. The differentiating elements of a strong bid are in the craft details: your specific references on comparable projects, your proprietary methodology, your certifications, your site organization. If you do not feed this information to the AI, it will not have it. Revision is not a formality — that is where the real value gets created.

Coverage limited to public construction contracts

Olra only covers public-sector contracts in the construction and building trades. Private tenders, IT procurement, white-collar service contracts, and architectural design competitions are out of scope. This is a strength (deep sector expertise) but also a limitation if your work is split between public and private.

Page volume caps apply

Plans are capped on pages processed per month (500 pages on Essential, 1,500 pages on Pro). For large tender packs or high-volume businesses, it is worth anticipating overages at €0.15 per additional page, or moving to the Enterprise plan.

No integration with construction business software yet

As of April 2026, Olra is a standalone web platform. There is no direct integration with common construction management software such as Batigest, Sage Bâtiment, or Obat. Data exchange is handled by file upload and export. Companies that want to integrate Olra into an automated workflow with their ERP will need custom connectors — possible, but requires specific development.

Our Field Assessment

After analyzing the platform and its market positioning, here is our synthesis.

Olra solves a real problem. Responding to public contracts is time-consuming, and construction tradespeople and SMEs who do it regularly are often handicapped by a lack of resources to process bids properly. Missing scoring criteria, overlooking risky contract clauses, mispricing relative to the market: these are common and costly mistakes. Olra addresses these with well-designed modules.

The tender-pack brief and proposal audit are the strongest features. Automatically extracting weighted scoring criteria and detecting the buyer's implicit expectations are real time savings that are difficult to replicate with generalist tools. The 14-point audit with score simulation is a genuine market differentiator.

AI writing is an aid, not a turnkey solution. Credit to the publisher for communicating this clearly. Be skeptical of tools that promise a perfect technical proposal in one click — public procurement is too specific for that. Olra's first draft saves time on structure and generic content, not on substance.

The ROI is fast for the right profile. At €149 excl. VAT/month, Olra pays for itself on a single contract won or a single poorly calibrated bid avoided. For a business responding to 2 to 5 tenders a month, the return on investment is straightforward. For occasional use, it is not justified.

Frequently Asked Questions