RFQ for AI software licenses (2-Year) – national lotteries commission

Closing date and time: 03 September 2026 @ 11:00. Bid advertisement date: 28 August 2026. No compulsory briefing session. Tender validity period: 90 days from closing date.

All questions regarding this RFQ must be forwarded to [email protected]. Enquiries will be responded to within 3 days of receipt. NLC contact details: Tel: +27(12)432 1300, Info Centre: 086 00 65383, Web: www.nlcsa.org.za, P.O Box 1556, Brooklyn Square 0083, Pretoria.

Submit the bid document via E-TENDER at https://www.etenders.gov.za/. No email or hard copies will be accepted. Late and incomplete submissions will not be accepted. Bids must be submitted in English. Returnable forms include SBD 1 (Invitation to tender), SBD 6.1 (Preference points claim), POPIA consent form, CSD summary registration report, valid SARS tax pin, B-BBEE certificate or sworn affidavit, SBD 4 (Declaration of Interest), OEM partner certification, JV agreement (if applicable), and the completed pricing schedule (Annexure A).

Evaluation is conducted in stages: Stage 1: Administrative compliance (pre-qualification) – must submit SBD 1, SBD 6.1, tax pin, CSD registration, POPIA consent, B-BBEE certificate. Stage 2: Mandatory compliance – SBD 4, OEM partner certification, JV agreement (if applicable), pricing schedule. Stage 3: Technical evaluation – rated on a 0-5 scale across criteria including company experience (20%), AI deployment experience (20%), methodology (30%), personnel experience (10%), and personnel qualifications (15%). Minimum threshold score is 70% for functional evaluation; only bidders scoring above 70% proceed. Stage 5: Pricing and specific goals – evaluated under the 80/20 preference point system, with 20 points for specific goals (e.g., Black Owned).

The successful bidder must supply eight (8) enterprise-grade user licenses for any two (2) of the following AI platforms: OpenAI ChatGPT Business/Enterprise, Azure AI Foundry/Azure OpenAI, Anthropic Claude Team/Enterprise, Google Vertex AI/Gemini API, or any other similar AI platform. The tool must include premium enterprise features: uncapped or high-priority access to advanced LLMs, expanded context windows and data analysis capabilities, dedicated workspaces, custom GPT/agent creation, and data sharing controls. Configuration must include an administrative control console, Single Sign-On (SSO) integration, domain verification, and data privacy assurance that prompts/uploads are not used to train public AI models. Technical support: SLA covering minimum 8 hours per month for troubleshooting and account maintenance, plus a dedicated helpdesk channel. Training: end-user training for 10 staff members and administrator-specific training for internal IT personnel. Deliverables timeline: Phase 1 – Project Kick-off & SCM Formalities within 1 week of appointment; Phase 2 – License Provisioning & SSO Integration within 2–3 weeks; Phase 3 – Admin & End-User Training within 4 weeks; Phase 4 – Ongoing Support & Maintenance for 24 months. Reporting to ICT Division. Contract duration: 2 years.

The methodology approach and execution plan is weighted at 30% of the technical evaluation. The bidder must submit a project implementation plan addressing the scope of work. The rating scale of 1-5 will be applied. A score of 0 is given for no project implementation plan. References for similar work completed within the last 5 years will be accepted, and multiple references for the same client may be regarded as one reference for evaluation purposes.

The bidder must submit a company profile of not more than 5 pages demonstrating capacity and clearly indicating the number of years in business providing software license deployments. Points for AI deployment experience: 5 points for above 5 years, 4 points for above 4 up to 5 years, 3 points for above 3 up to 4 years, 2 points for above 2 up to 3 years, 1 point for above 1 up to 2 years, 0 points for less than 1 year. The Lead Technical Expert must submit an updated CV including relevant qualifications and at least three traceable references demonstrating experience in deploying/supporting AI tools, with valid certification from the software vendor. Personnel experience is weighted at 10%.

Personnel qualifications: The Service Manager / Lead Technical Expert must hold a valid certification from the software vendor. Points are awarded based on years of experience with valid certification: 5 points for more than 5 years, 4 points for more than 4 up to 5 years, 3 points for more than 3 up to 4 years, 2 points for more than 2 up to 3 years, 1 point for more than 1 up to 2 years, 0 points for less than 1 year or no CV with valid certification. Qualifications: 5 points for postgraduate qualification (Honours, Master's or higher) in ICT or relevant field, 4 points for Bachelor's Degree, 3 points for National Diploma, 2 points for Diploma, 1 point for Certificate, 0 points for no qualification or not relevant. All qualifications must be verifiable through SAQA or other recognised body. Minimum threshold score for functional evaluation is 70%.

The pricing schedule is provided in Annexure B: PRICING SCHEDULE TEMPLATE. Bidders must complete the pricing schedule according to the NLC Pricing Schedule (Annexure A) and submit it with their proposal. The evaluation for pricing and specific goals will include the pricing schedule.

Pricing must be submitted according to the NLC Pricing Schedule (Annexure A). The RFQ will be evaluated under the 80/20 preference point system. No other financial requirements (bonds, guarantees, payment terms) are specified in the document.

CSD registration is mandatory (except foreign suppliers with no local registered entity). Tax compliance confirmation via valid SARS Tax Pin. B-BBEE certificate or sworn affidavit required (valid B-BBEE certificate in terms of Codes of Good Practice, or Exempted Micro Enterprises Sworn Affidavit, or Consolidated B-BBEE/Sworn Affidavit for JV). SBD 4 (Declaration of Interest) must be fully completed and signed. OEM status: bidder must hold valid accredited partner or reseller status with the OEM for any two of the four AI platforms. POPIA consent form required. If JV, signed JV Agreement/MoU, authorized representative, company registration & statutory documents, and consolidated B-BBEE/Sworn Affidavit required. Bidders must sign and submit the Standard Bidding Document (SBD).

The bidder warrants that all its employees (including sub-contractors) are covered in terms of the Compensation for Occupational Injuries and Diseases Act, and that the cover remains in force for the duration of the adjudication and any subsequent agreement. The NLC reserves the right to request documentary proof of registration and 'good standing' with the Compensation Fund.

The contract duration is 2 years, effective on the date of appointment, date of signing of an SLA, or as directed by the NLC. No services must be rendered or goods delivered before an official NLC Purchase Order form has been received. The bidder warrants that all employees (including sub-contractors) are covered under the Compensation for Occupational Injuries and Diseases Act. The NLC conducts regular performance reviews, with a minimum of an annual review for contracts longer than a year. Non-performance will be addressed with a formal letter specifying remedial actions; non-adherence may lead to escalating performance management actions. The bidder must provide an SLA covering minimum 8 hours per month for troubleshooting and account maintenance, and a dedicated helpdesk channel. The bidder must ensure data privacy: all organizational prompts, uploads, and data remain confidential and are not used to train public AI models. Bidders must not make any news releases concerning the RFQ without NLC consent. The RFQ document is confidential and its contents may not be used or disclosed except for preparing a proposal.

The RFQ will be evaluated under the 80/20 preference point system. Specific goals are allocated points for procurement from entities who are Black Owned, with a maximum of 20 points for specific goals under the 80/20 system. The technical evaluation uses a rating scale of 0-5 (Excellent=5, Good=4, Acceptable=3, Minor Reservations=2, Serious Reservations=1, Unacceptable=0). The technical/functional evaluation includes: Company Experience (20%), AI Deployment Experience (20%), Methodology Approach and Execution Plan (30%), Personnel Experience (10%), and Personnel Qualifications (15%). Minimum threshold score is 70% for functional evaluation; only bidders scoring above 70% proceed to pricing and specific goals evaluation.