Human-centered design systems and product experiences
Core services
01UI/UX Design02Web Development03Software Development042D/3D Experiences05App Development06AI Agents & Automation07Website DesignTechnical services
Frontend Engineering
React DevelopmentNext.js DevelopmentWeb ApplicationsEnterprise DashboardsSaaS PlatformsBackend & Infrastructure
API DevelopmentCloud InfrastructureDevOps & CI/CDBackend SystemsCustom SoftwareIndustries we serve
Technology & Software
AI Prospect Research Agent Development
The account, read before the call
An agent that does the reading before a sales conversation.
An agent that does the reading before a sales conversation. It researches the accounts you choose from sources you approved, and returns a short brief in which every fact carries its link and the date it was read.
Good account research is slow, and bad research is worse than none. A rep spends forty minutes across a website, a news search and a registry, or five minutes with a tool that returns a confident paragraph nobody can verify. Then a meeting opens with a fact that turns out to be about a different company with a similar name.
This page describes a workflow we propose to build as a custom project. It does not exist as software yet. It would build on the account research that runs in our outreach preparation demonstration on test data, where each fact keeps its link and the date it was read. Contact research, fuller briefs and a research queue are not built.
Six jobs it carries through
Source-bound research
Reads only the sources you approved.
- Company website and newsroom
- Registries and filings you list
- Licensed data providers
- Nothing from sources off your list
Company identification
Makes sure it is the right company.
- Matches on domain and registration
- Notes group and subsidiary structure
- Flags similar names
- Asks when identity is unclear
Cited facts
Records each fact with its evidence.
- The statement as the source wrote it
- The link and the date read
- The date of the source itself
- A note when sources disagree
Relevance to your offer
Says why a fact matters to you.
- Uses the criteria you sell on
- Marks facts that change the conversation
- Separates fact from inference
- Lists what it looked for and did not find
People research
Identifies relevant roles, within your rules.
- Roles from sources you approved
- Where each detail came from
- Do-not-contact marks respected
- No scraping of personal profiles
Account brief
Delivers a brief your team can use.
- One page per account
- Facts, relevance and open questions
- Stored on the account record
- Refreshed when you ask
How the Work Is Done Today
Facts without a source
A brief says the company is expanding and nobody knows where that came from.
The wrong company
Two firms share a name, and the brief mixes them.
Stale information
A funding round from three years ago is presented as news.
Contacts of unknown origin
Names and emails appear with no record of how they were obtained.
Everything, unranked
Thirty facts, none marked as mattering for your offer.
Research nobody reuses
Each rep researches the same account again.
How One Request Moves Through
An example of one case, in 6 steps. The same steps play in the picture at the top of this page.
Account chosen
A rep adds a logistics company to the research queue before a first call.
Company confirmed
The domain and the registry entry agree. A similarly named haulier is noted and set aside.
Sources read
The company's own site and newsroom are read, along with the registry you approved.
Facts recorded
A new depot announced last month is recorded with its link, the date read and the article date.
Relevance marked
The depot is marked as relevant because your criteria include multi-site operations.
Brief delivered
A one-page brief lands on the account record, with two questions it could not answer.
Walk One Real Case Through It
Thirty minutes, one case from your own process. We map it against the workflow on this page and tell you plainly what can be reused from our demonstrations and what has to be built for your systems.
Where It Connects
The systems named here are the kind this agent works with. Which ones it connects to in your case is settled in the scope.
Research sources
- Company websites and newsrooms
- registries
- licensed data providers
CRM
- Account and contact records
Criteria
- What makes an account relevant to you
Preferences
- Do-not-contact and suppression records
Research store
- Where facts, links and dates are kept
Where reps work
- CRM, email or chat
Every Fact Has a Link and a Date
The brief is only as good as its evidence, so the agent keeps the source with each statement and labels anything it inferred. If two sources disagree it shows both. If it found nothing, it says what it searched.
It reads the sources on your list and no others. Whether a source may be used, and what personal data you may hold about a contact, are decisions for your business.
- Only sources you approved
- A link and a date for every fact
- Inference labelled as such
- Wrong-company risks flagged
- Personal data within your rules
- It contacts no one
Asked Before the First Call
How is this different from the AI SDR and outreach agent?
Research is one step of outreach, and here it is the whole job. This agent produces briefs for people: before a call, a meeting or a proposal. It does not screen contacts for outreach, draft messages or handle replies.
How is this different from the competitor intelligence agent?
Prospect research looks at one company you want to sell to, on request. Competitor intelligence watches a fixed set of companies you compete with, continuously, and reports changes. The source rules are the same.
Where does the information come from?
From sources you approve in advance. Typically that is a company's own website and newsroom, public registries and any data provider you hold a licence for. The agent does not scrape personal social profiles, and it does not use a source that is not on your list.
What exists today?
Not as software yet. This page describes a proposed workflow. It would build on the account research that runs in our outreach preparation demonstration on test data, where each fact keeps its link and the date it was read. Contact research, fuller briefs and a research queue are not built. A project starts by mapping your process, then proving the first steps on test data before anything touches a live system.
Pick Five Accounts You Know Well
We will research them and you can check the briefs against what you already know. That is the fastest way to see what is missing.