AI Industry & Applications · 2026-03-20 · 05:00
HSC Pipeline Engineering: building an engineering knowledge base with RAG AI
In Brief
Through the AISG LADP programme, HSC Pipeline built a locally deployed RAG AI knowledge base, breaking down engineering-knowledge silos and improving decision-making efficiency.
Key Takeaways
- HSC Pipeline lays gas, water, district cooling, power and telecom utilities; site queries used to wait 3 days for a senior PM to respond.
- Through AISG's LADP, HSC built an on-premise RAG AI assistant accessible only over the company intranet.
- The AI Playbook returns accurate answers with cited sources, cutting delays that previously triggered complaints, demerit points and lost tenders.
Summary
HSC Pipeline is a civil engineering firm laying utilities for the nation. Field staff face a steady stream of questions every day, but clashing schedules meant even minor issues could wait three days for a project manager's response. Backlogs snowballed into delays, which drew public complaints, demerit points from clients, and threats to future tenders.
Through AI Singapore's LADP, HSC built an on-premise RAG-enabled AI assistant called the AI Playbook, accessible only over the company intranet. It reads internal documents and answers staff queries with cited sources, so teams can verify the basis of every recommendation. Project teams stopped hunting for answers and started resolving issues directly. HSC also uses the tool as a talking point in its internal AI literacy campaign — a way to teach staff about confidentiality and protection of sensitive documents. The company recommends other construction firms join LADP to start their own AI work.
Full transcript
Caption language: en · Fetched: 2026-05-02
HSC is a civil engineering company that does infrastructure for the nation. We provide services to lay various types of utilities like gas, water, district cooling, power cables and telecom. Our staff actually face a large variety of problems on a day-to-day basis. Of course, we have the option of uh asking the more senior project managers or the project managers on what is the resolutions to such issue. But due to clashing schedules, this often can take up to 3 days to resolve even a minor issue. So when issues start to snowball, this have larger implications to the project. Complaints from the public will actually damage HSC's reputation as a reliable civil engineering company. If the field tasks are not completed, they will snowballs into delay and causing the project timeline to sleep.
Clients then will impose demerit points or fines and this will affect our ability to complete the current projects and also our future tenders. Through the LAP, we built a secure onremise RAG enabled AI assistant that is only accessed through the company's internet. It will take in our documents and use this context to provide funline support to staff queries. It gives accurate enough answers and site sources for traceability back to what was used to arrive at these responses. The AI playbook brings up the most relevant document pages for any scenario at hand. There is less ambiguity regarding whether a cost of action is applicable in the situation. Project teams spend a lot less time looking for answers and instead can refocus to quickly and accurately resolve issues.
This has translated to significant time and cost savings for our timelines. We've tested the AI playbook with our project teams and the response has been largely positive. We used the AI playbook as a talking point in our company's internal AI literacy campaign. We used it to emphasize the importance of business confidentiality and the steps to take to protect our trade secrets, sensitive information and business critical documents. HC is a construction company. We lack expertise on how to design, build and manage AI tools. AISG's LAP provided training and consultation to overcome this shortfalls. Furthermore, AISG's LAP gave us access to professional Gen AI models and case studies. This enabled us to quickly build and test prototypes that eventually led to the AI playbook.
When the playbook was able to consistently and accurately answer our questions on any projects, that was the moment AI changed from a lofty aspiration for us to something more concrete and attainable. Our experience working with the AISG team has been very good. They are well versed in applying AI to use cases and extracting business value from it. Whenever we had questions or stumbled in our app development, they have been supportive and provided much needed knowledge on how to address our various challenges. ESG's LAP is a cost-effective program to quickly pick up the basics of AI app development and drive business value from lowhanging fruits in any company on the front of digital automation. HSC highly recommends other companies to register for the program to jumpstart their AI journeys.
Related Videos
Ong Ye Kung on AI, genetic screening and preparing for a super-aged Singapore
2026-03-04 · Ong Ye Kung · 30:36
Health Minister Ong Ye Kung talks through AI applications in healthcare and Singapore's strategy for a super-aged society.
YTL PowerSeraya: LLMs power electricity market rule analysis
2026-02-20 · YTL PowerSeraya · 05:00
Singapore power company YTL PowerSeraya used LADP to build an LLM specialised in electricity market rules, enabling automated report analysis and rule queries.
Skybots: from RPA to LLM-powered customer service
2026-01-15 · Skybots · 05:00
Accounting-tech firm Skybots used LADP to upgrade RPA into LLM-powered customer service, handling complex accounting workflow queries.
Josephine Teo on AI's role in SMEs, education and society
2025-11-19 · Josephine Teo · 06:17
Josephine Teo on how AI can help SMEs transform, reshape education, and reach every part of society.
Josephine Teo on how AI uplifts Singapore's financial services industry
2025-10-06 · Josephine Teo · 09:49
Josephine Teo discusses prospects for AI applications in Singapore's financial services and the regulatory balance required.
Multispectral AI transforms plastic waste recycling and sorting
2025-07-01 · AI Singapore · 03:00
An AISG 100E project applies multispectral AI to convert plastic-waste sorting from a manual process to an automated one, addressing a long-standing pain point in the recycling industry.