# AI lead generation and AI SDR > We use AI to do the slow parts of outbound research: spotting a buying signal, reading up on the account, finding the right role and drafting a short message tied to that signal. A person approves every message before it goes, and replies go to your sales team. Source: https://www.fyadvisers.com/services/ai-lead-generation/ Author: FYAdvisers Last updated: 2026-10-13 AI lead generation means using AI to find companies with a reason to buy now, research them and draft the first message, so your salespeople spend their time on conversations rather than lists. FYAdvisers runs this for B2B companies selling to Singapore businesses, and for overseas teams selling into Singapore. AI watches for buying signals, reads up on each account and drafts outreach tied to the signal. A person approves every message before it is sent, and replies go straight to your sales team. This is the AI side of our [signal-based lead generation](/services/signal-based-lead-generation/) service. The approach is the same: fewer messages, each with a stated reason to write. ## What an AI SDR does, and what we let it do Most AI SDR and AI sales agent products promise to do the whole job of a sales development representative: find contacts, write messages, send sequences and book meetings. We split that job in two. **What AI does well:** - **Spotting signals.** Reading tender awards, job adverts, company announcements and news, and flagging the ones that match your customer profile. - **Account research.** Summarising what a company does, what changed and why your offer might be relevant, from public sources. - **Finding the role.** Suggesting which job title is most likely to own the problem, so the message reaches the right desk. - **Drafting.** Writing a short first message that refers to the specific event, in your tone, for a person to edit. - **Sorting replies.** Labelling replies as interested, not now, wrong person or unsubscribe, so your team sees the important ones first. **What stays with people:** - Approving every message before it is sent. - Deciding who is contacted and through which channel. - Replying to prospects, quoting and making any commitment. An AI that sends messages on its own can write something wrong, irrelevant or awkward under your name at scale. A person reading each draft catches that before a prospect does. ## Typical uses for Singapore B2B companies - **Contract and tender awards.** A company that has just won a contract may need suppliers, staff or services to deliver it. AI can read award notices and draft a note to the right person. - **Hiring signals.** A job advert for a role you support suggests a new team or a gap. AI can group adverts by company and summarise what they imply. - **Expansion and new sites.** New offices, warehouses or outlets often bring fit-out, IT, cleaning and logistics needs. - **Inbound enquiries.** An enquiry to your own business is the strongest signal of all. Our [speed-to-lead](/services/speed-to-lead/) work uses AI to summarise and route enquiries so a person can reply quickly. ## How a project runs 1. **Discovery.** We look at deals you have won, who bought and what had just happened at their company. That tells us which signals to watch and who to contact. 2. **Prototype with your real data.** We run the research and drafting on a small set of real accounts and show you the output. You mark what is useful and what is not. 3. **Build.** We set up the signal sources, research steps, drafting instructions and an approval queue, and connect them to your CRM or a shared sheet. See [CRM implementation](/services/crm-implementation/) if you do not have one yet. 4. **Human review.** Every draft waits for approval. You can edit, reject or send. 5. **Monitoring.** We log which signals, accounts and messages lead to replies, review drafts that were rejected, and adjust the instructions. ## What it costs We outline our own fees after a conversation, because they depend on how many signals you want watched and how many markets you cover. The main cost drivers are: - **Signal sources.** Some are free public records. Others are paid data feeds or databases. - **Contact data.** Verified business emails and phone numbers usually come from a paid provider. - **Model usage.** AI providers bill per token, so cost rises with the number of accounts researched and the length of each summary. - **Review time.** Someone on your team, or ours, reads every draft. For comparison, if you want to run tooling yourself: Clay's pricing page listed a free tier and a Launch plan from $167 a month billed annually ($185 billed monthly) when we checked in October 2026. Apollo's pricing page lists Free, Basic, Professional and Custom plans. Check the vendor pages for current prices. ## Data, privacy and outreach rules Outreach written with AI follows the same rules as outreach written by a person. - **Do Not Call.** PDPC's [Advisory Guidelines on the Do Not Call Provisions](https://www.pdpc.gov.sg/-/media/files/pdpc/pdf-files/advisory-guidelines/advisory-guidelines-on-the-dnc-provisions-1-feb-2021.pdf) (revised 1 February 2021) exclude messages sent to an organisation for the receiving organisation's purposes, which covers genuine business-to-business marketing. Marketing to someone in their personal or domestic capacity is not covered by that exclusion, and calls or texts to Singapore numbers then need a DNC Registry check unless you hold the right consent. PDPC's [guide for businesses](https://www.pdpc.gov.sg/overview-of-pdpa/do-not-call-registry/business-owner/do-not-call-registry-and-your-business) explains the registry. - **AI and personal data.** PDPC's [Advisory Guidelines on use of Personal Data in AI Recommendation and Decision Systems](https://www.pdpc.gov.sg/-/media/files/pdpc/pdf-files/advisory-guidelines/advisory-guidelines-on-the-use-of-personal-data-in-ai-recommendation-and-decision-systems.pdf), issued on 1 March 2024, explain how the PDPA's consent, notification and accountability obligations apply when AI systems use personal data. - **Governance.** IMDA and the AI Verify Foundation released the [Model AI Governance Framework for Generative AI](https://aiverifyfoundation.sg/resources/mgf-gen-ai/) on 30 May 2024. Its dimensions, including accountability, data, and testing and assurance, are a useful checklist for any AI outreach setup. In practice, we research companies from public sources, contact people in their business role, refer only to what is public or what the prospect told you, and keep a record of approvals. This is general information, not legal advice. ## Build or buy **Buy an AI SDR tool** if you have someone who will own it: choosing data, writing prompts, checking output and managing sending limits. Clay, Apollo, HubSpot, 11x and Artisan are well-known options with different strengths. **Work with us** if you want the signals chosen, the research done and drafts ready to approve, without hiring or training someone to run the tooling. We can also set up the tools for your team to run later. **Do neither** if your market is a few dozen named accounts. A salesperson with a good list and a calendar may do better than any automation. ## Who this suits - B2B companies with a clear customer profile and a deal size that justifies research per account. - Teams whose salespeople are spending their time building lists instead of talking to buyers. - Overseas B2B teams selling into Singapore that want local signals and context. It suits consumer businesses less, because personal-capacity marketing brings stricter consent and Do Not Call requirements. ## What to prepare - Three to five recent deals you won, and what was happening at each customer at the time. - The job titles that usually buy from you. - The channels you are comfortable using, and any rules about who must not be contacted. - Your CRM or the sheet where replies should land. - Who on your team will approve drafts, and how quickly they can do it. [Get in touch](/#contact) and we will suggest the signals we would start with. ## Frequently asked questions ### What is an AI SDR? An AI SDR (sales development representative) is software that does some of the work a junior salesperson does before a meeting: researching companies, finding contacts, writing first messages and sorting replies. Some products send messages automatically. We use AI for the research and drafting, and keep a person approving what is sent. ### Is AI lead generation the same as an AI sales agent? The terms overlap. An AI sales agent usually means software that takes actions on its own, such as sending emails or booking meetings. AI lead generation is broader and can stop at research and drafts. Our service stops at drafts a person approves, and replies are handled by your team. ### Do the PDPA and Do Not Call rules apply to AI outreach? Yes, in the same way they apply to outreach written by a person. PDPC's Do Not Call guidelines exclude marketing messages sent to an organisation for the organisation's own purposes, but not marketing to someone in their personal capacity. We agree channels and contact rules with you before anything is sent. This is general information, not legal advice. ### Do you guarantee meetings or a number of leads? No. We do not promise reply rates, meetings or pipeline. We agree what to track at the start and show you which signals and messages lead to replies, so you can judge the work yourself. ### Can we use an AI SDR tool instead? Yes, and for some teams that is the better choice. Tools such as Clay, Apollo, 11x and Artisan cover research, data and sequencing. Our service suits teams that want the signal selection, research and drafting done for them, with a person checking every message.