Account research

Account research is how Enki IQ turns a company name into a briefing you can walk into a meeting with. Every account in your territory can be researched, and the depth of the research depends on the tier. This article covers what the pipeline produces, how T1 differs from T2 and T3, and what you see on the account when the work finishes.

What you get

Every completed research run produces the same six sections on the account: Overview (what the company does, where they operate, recent moves), Financials (revenue, growth, funding, public filings if relevant), Strategy (priorities, initiatives, stated direction), Market (the market they play in and how they're positioned), People (leadership, key teams, hiring trends), and Sales (who buys, how they buy, the shape of the sales motion). A Sources list captures the URLs the research drew from so you can verify any claim.

The sections appear on the account's Research tab. You can read them inline or export the whole briefing to PDF. Once research lands, it also feeds the rest of the product: foresights, competitive intelligence, talking points, the territory plan, and the practice persona generator all draw on the saved research.

Tier 1: Deep Research

Tier 1 accounts get Enki IQ's highest-quality research path: Deep Research. Deep Research uses a hosted agent that browses the web, reads source materials, and synthesizes a comprehensive briefing across all six sections. It's the most expensive option and the slowest, but the depth is meaningfully higher: it can chase down a footnote in an annual report or trace a hiring spree across multiple sources.

Deep Research is asynchronous. When you trigger it from the account page, Enki IQ acknowledges the request, the job runs on a background worker, and a progress loader on the page shows you how far along it is. You can leave the page, refresh, or come back tomorrow; the result lands when ready. Deep Research is metered against your monthly Account Deep Research quota.

Tier 2 and Tier 3: the standard pipeline

Tier 2 and Tier 3 accounts go through Enki IQ's standard research pipeline. It produces the same six sections as Deep Research, faster and at lower cost, by breaking the work into five clear stages. You don't see these stages individually in the UI — they're the internal pipeline names — but understanding the structure helps you read the result.

PLAN is the first stage. Enki IQ decides what specifically needs to be looked up about this account given what's already known. The plan is a structured list of questions and the sources most likely to answer them.

SCRAPE is the second stage. Enki IQ fetches the pages and documents identified in the plan via BrightData, which handles the actual web retrieval against the sources.

SYNTHESIZE is the third stage. Enki IQ reads the scraped material and reasons over it: extracting facts, weighing conflicting claims, and assembling a draft briefing. Gemini Pro does the heavy reasoning here.

FORMAT is the fourth stage. The synthesized draft is parsed into the structured shape the UI consumes — the six named sections with their typed fields. Claude Sonnet does the formatting, and the formatter is the only place that produces the final shape of what you read on screen.

SAVE is the final stage. The formatted research is persisted to the account, semantic embeddings are generated and stored so the new content is searchable, and the account's Research tab updates. After SAVE, every downstream surface that reads from this account (foresights, talking points, gap analysis) can use the new material.

Like Deep Research, the standard pipeline is asynchronous and surfaces progress through the same loader. Standard pipeline runs and Deep Research runs each consume their own monthly quota line, so a month's worth of standard research doesn't draw down your Deep Research budget and vice versa.

When each pipeline runs

Deep Research runs on Tier 1 accounts and on any account where you explicitly pick it from the research tier dialog on the account page. The standard pipeline runs on Tier 2 and Tier 3 accounts by default, including the cohort runs the territory worker kicks off when building or refreshing your territory plan.

In practice this means: if you've just imported an account and confirmed it as Tier 1, expect Deep Research depth. If you've imported a long tail of Tier 2 and Tier 3 accounts and want them ready before next quarter, the standard pipeline gets you a respectable briefing on each for a fraction of the cost.

What changes on the account when research lands

When a research job completes, the Research tab on the account fills in. The Overview tab summarizes the headline facts in a glanceable card. Foresight signals may start appearing on the account if Enki IQ detects buying signals in the new material (see "Foresight"). Competitor mentions in the research feed the Competitors tab, and the gap analysis surface becomes meaningful as soon as at least one competitor is identified (see "Competitive intelligence"). Talking-points cards and the per-account briefings on the Territory page draw on the new research the next time they refresh (see "Territory and scoring").

Clicking Generate twice on the same account doesn't start two jobs. The second click adopts the in-flight job and shows you the same progress loader. If you refresh the page mid-run, Enki IQ finds the running job and reattaches.

Under the hood

Deep Research runs Gemini's hosted research agent. The standard pipeline composes Gemini Pro for the plan and synthesis stages, BrightData for the scrape, and Claude Sonnet via Vertex for the format stage. Voyage AI embeddings are generated on save. Failures in one stage don't require redoing the work that already succeeded.