T
Trakti
Trakti is a contract execution platform for legal, procurement, sales and operations teams, made by Trakti Ltd of London. Teams turn approved language into templates, negotiate in a shared room, sign electronically and track obligations and renewals after signature. Advanced and Corporate plans add blockchain notarization and smart contracts that carry out agreed steps such as milestone payments or refunds.
Gemini models run Model Extraction, Deal Extraction and Contract Extraction, which turn an uploaded agreement into a template, a deal record or a live contract record. Agents suggest clauses against the playbook, draft from templates, propose redlines and prepare counter offers. Every AI action is logged with the model, its confidence level and the approver, and teams set escalation thresholds for when agents defer to legal. Salesforce, HubSpot, Dynamics 365, SAP, Coupa and Microsoft 365 connections move deal data in and signed terms back.
Trenitalia runs its Smart Refund delay refunds on Trakti, and Leonardo built DEServE for service contracts, which won a World Commerce and Contracting award in 2022. Lite starts at €34 per user a month billed annually, Advanced is €159 with at least ten users, and Corporate is quoted after a 14 day trial.
Alternatives to Trakti
Other products that do the same kind of work as Trakti, and what each one does.
- LLawliftLawlift is a document automation platform for law firms and corporate legal departments. Lawyers turn existing agreements into approved templates with questionnaires, conditions, calculations, and reusable clauses.
- SSmarter Drafter ProSmarter Drafter Pro turns firm precedents into guided drafting workflows for law firms and in house legal teams. Tensis Group Pty Ltd builds it around a form builder, a clause bank in Microsoft Word and a rules engine that assembles documents from approved wording.
- AAvokaadoAvokaado is a contract lifecycle platform from Avokaado.com OÜ of Tallinn, Estonia, founded by Mariana Hagström, who leads it with cofounder Jean Mauris.
- GGitLawGitLaw is contract drafting and review software for small businesses, founders and lawyers. Nick Holzherr founded the company in 2025, and GitLaw Inc. operates from San Francisco.
- LLeegalityLeegality is an Indian document execution and contract operations platform operated by Grey Swift Private Limited. It combines Aadhaar and other electronic signatures with digital stamping, internal approvals, a contract repository and obligation reminders.
- BBigleBigle, formerly Bigle Legal, combines contract lifecycle management with Libra, an AI assistant for legal departments and the business teams that work with them.
Capability grades
All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
Gemini models run Model Extraction, Deal Extraction and Contract Extraction, which turn an uploaded agreement into a template with variables, a deal record or a live contract record, and agents suggest clauses, draft from templates, propose redlines and prepare counter offers on standard terms. Templates, negotiation rooms, eSignature, obligation tracking and smart contract rules that trigger payments or refunds all run without the models. The AI speeds the work around a contract execution platform that stands on its own.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Extraction works from the uploaded agreement and is built to leave renewals, terminations, obligations and classifications blank where the text does not support them. Retrieval and semantic reranking of contract excerpts feed the AI responses, and complex formatting can yield a partial or simplified result. The AI terms warn that output may contain inaccuracies, omissions or misclassifications. No accuracy figure is given, and output does not link each extracted value back to its passage.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
Agents draft, redline and prepare responses for the legal team to approve, and every AI decision can be reviewed, approved or overridden, with nothing changed without approval. Teams set the approved clause library, acceptable deviation ranges and mandatory terms, and configure escalation thresholds that tell agents when to defer to legal. Administrators control which users reach AI functions. Each AI action is logged with the model that made the suggestion, its confidence level and the approver, and the AI terms leave decisions with legal effect to people.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
Trenitalia launched Smart Refund in March 2025, using Trakti's contract logic to pay delay refunds under the carrier's terms. Leonardo's services organization built DEServE on Trakti to manage customer and supplier service contracts, and that project won the World Commerce and Contracting President's Award in 2022. Neither account gives figures, and neither concerns the current AI extraction and agent features.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
Clause 4.1 of the AI terms states that customer contract data is not used to train shared AI models unless explicitly agreed, and the processing agreement requires prior explicit written consent. Contracts, templates and workgroups carry role based visibility, and AI processing runs under a zero data retention configuration. Gemini may cache request data in encrypted, project partitioned storage for up to 24 hours. Privilege and work product are not addressed.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
The AI terms state that AI output is not legal advice and may contain inaccuracies, omissions or misclassifications, and that it does not replace professional legal review. Users remain solely responsible for reviewing, validating and approving AI output before use, and Trakti recommends review by a legal professional before anyone relies on it. Jurisdiction limits and supervision of the sales and procurement users who work in the platform are not covered.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
Administrators assign AI roles and can see who uses which capability, and the agent audit trail records the model, its confidence level and the approver for each action. Teams encode their playbook as constraints the agents must follow, and the security program refers to impact assessments for higher risk processing and internal testing of AI workflows. No results on uneven performance by contract type, language or jurisdiction are given, and no accountable owner is named.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
The data processing agreement requires breach notice within 48 hours of awareness, though the general security overview speaks of 72 hours. Its Schedule 2 points to the subprocessor list, with AWS in the EU, Cloudflare, Google Cloud with Gemini and AI Studio in the EU and EEA, SendGrid, HubSpot, Stripe and others named with locations, and customers can object to a new one and terminate without penalty. Schedule 3 covers TLS 1.2 and AES 256 encryption, MFA for administrators, intrusion detection, logging and tested restores.
AI processing keeps nothing after extraction beyond Gemini's 24 hour cache, while contract data is processed for the life of the terms and any further period the law requires, with no fixed deletion date.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
Under the AI terms Trakti is not liable for decisions made on AI output, the terms of use disclaim liability for inaccurate extracted data, and the AI terms cap liability at the amounts in the subscription or service agreement. The general terms refuse any guarantee of specific results or continuity of service. English law governs, and no remedy covers a wrong extraction or a bad suggested clause.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
Salesforce, HubSpot and Dynamics 365 deals prefill a contract with client, product and price data, and signed terms, values and renewal dates flow back to the pipeline. SAP, including Ariba, receives purchase orders, supplier records and obligations, Coupa is listed, Microsoft 365 receives signed copies and attachments, and milestones can trigger invoices, payments or refunds in accounting systems such as Xero and Sage. Setup is described only as connecting a native connector or the REST API, followed by prefill, execution and sync back.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
Contract data is hosted on AWS in the EU, and Gemini document analysis, retrieval and reranking run on Google Cloud in the EU and EEA, so processing and storage are each placed. Temporary data sent for AI processing is held apart from saved contract records. Support, email and payment providers such as Cloudflare, SendGrid, tawk.to, Stripe and PayPal sit in the United States, and no regional choice beyond the EU is offered.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
Trakti holds ISO 27001 certification to the 2022 standard with annual surveillance audits, and the security program covers risk assessment, asset management, encryption and continuity. The certificate and a penetration test summary are released on request through the trust center, so the scope and dates cannot be read in advance.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
Google's Gemini Enterprise Agent Platform supplies the Gemini models, retrieval and semantic reranking for contract analysis, and Google AI Studio is listed for development and troubleshooting that can involve customer excerpts. Changes reach customers through updates to the subprocessor schedule, with a right to object. The specific Gemini model and version are not named.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
Lite is €34 per user a month billed annually or €39 monthly, Essential €69 or €79 with at least three users, Pro €99 or €119 with at least five and Advanced €159 or €179 with at least ten, all before VAT. The plans include 20, 40, 80 and 160 deals a month and 2 to 20 GB of storage, and Advanced adds blockchain notarization and smart contracts. Corporate, with unlimited deals, custom integrations and an SLA, is quoted. Which AI extraction features come with each plan is not shown, and a 14 day trial precedes the choice.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
Legal departments use Trakti to control approved language, review deviations and track obligations after signature, while sales, procurement, supply chain, HR and operations run their own workflows. Named deployments cover rail refunds at Trenitalia and service contracts at Leonardo, and smart contracts carry milestone payments and refunds. Law firms are not a stated market, and litigation and court work fall outside the product.
5 public documents
The public pages on file for Trakti, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.
-
trakti.com/terms6 signals
Client Data in Training, Prompt and Output Retention, Third Party Request and Subpoena Notice and 3 more
Read Nov 21, 2026
-
Fabricated Citation Record
Read Nov 21, 2026
-
trakti.com/ai-agents1 signal
Court Disclosure Support
Read Nov 21, 2026
-
trakti.com/integrations1 signal
Billing and Fee Posture
Read Nov 21, 2026
-
trakti.com/security1 signal
Ethical Walls and Matter Segregation
Read Nov 21, 2026
From €34 per user per month billed annually
- Lite costs €34 per user a month billed annually or €39 billed monthly.
- Essential is €69 per user with at least three users, and Pro is €99 with at least five.
- Advanced is €159 per user with at least ten, and Corporate is quoted.
- The trial lasts 14 days.
Plan rates exclude VAT, and monthly billing costs €39, €79, €119 and €179 per user for the four listed plans. Lite includes 20 deals a month and 2 GB of storage, Essential 40 and 5 GB, Pro 80 and 10 GB and Advanced 160 and 20 GB. Advanced and Corporate add blockchain notarization and smart contracts, and Corporate adds unlimited deals, custom integrations, an SLA and a dedicated account manager. All plans include a contract repository, audit trail and eSignatures.
Implementation: Plans include onboarding. Corporate setup and any related charges are set in its quote.
Note: Paid plans activate on payment, annual rates need the annual commitment and VAT may apply. Which AI extraction and agent features each plan includes is not shown on the price list.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
Training occurs only where the customer has affirmatively enabled it.
Customer contract data is not used to train shared or general AI models without the customer's prior explicit written consent, under the AI terms and the processing agreement.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
The customer sets the retention window and no retention is an available setting.
AI processing runs under a zero data retention configuration, so nothing is stored after extraction except Gemini's encrypted cache of up to 24 hours. Saved contracts, templates and operational logs follow their own retention.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Segregation is asserted in public materials with no published detail on how it is enforced.
Administrators set access by role down to individual platform components, and templates and workgroups carry their own controls, with no wall described for privileged matters.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
Trakti will tell the customer before any legally required processing beyond its instructions, unless the law prohibits that notice.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
No located public material identifies the corpus behind the product’s answers.
The AI retrieves and reranks excerpts from the customer's own contracts and approved clause library, with no collection of primary legal authority behind it.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
Trakti tracks versions, signatures and obligations inside agreements, and no court decision is cited that would need a good law check.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
The vendor describes refusal or abstention behavior in public materials.
Extraction leaves renewals, terminations, obligations and classifications unset where the contract text does not support them, and agents defer to the legal team at the escalation thresholds a team configures.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.
Damien Charlotin's AI Hallucination Cases database lists no court decision involving fabricated citations from Trakti.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Public materials refer to professional responsibility in general terms without naming guidance.
The AI terms separate contract extraction from legal advice and recommend review by a legal professional before reliance, without naming a bar opinion or practice rule.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The product does not touch a fee between a lawyer and a client. It operates before an engagement exists, or it is bought by a team that bills no client for the work. Savings claims aimed at the buyer’s own cost are recorded in the summary and do not make the row a savings claim, because no client bill is in the loop.
Buyers are company teams drafting, negotiating and executing their own agreements, so the platform's time savings and smart contract payments sit outside any client fee.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
A current subprocessor or model provider list is published.
The privacy policy names each provider with its location, among them AWS in the EU, Google Cloud and Gemini in the EU and EEA, Cloudflare, SendGrid and Stripe in the United States and Microsoft in the UK, and customers may object to a new subprocessor.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Some elements of the record are available, short of a document level export.
Each AI action is logged with the model, its confidence level and the approver, and contracts carry timestamps and signature proof, with optional blockchain notarization, though no export is shaped for a court disclosure.