Your always-on AI research analyst — delegate research, synthesis, and execution directly inside AlphaSense
Note: SuperAnalyst is currently in beta. Features, behavior, and screens described here may not yet be available or change before general availability.
On this page:
What is SuperAnalyst?
SuperAnalyst turns AlphaSense from a search-and-answer tool into a true end-to-end workflow engine. Instead of running searches and building the deliverable yourself afterward, you can hand the entire workflow, research, synthesis, analysis, and the final deliverable directly to your agent.
Think of SuperAnalyst as a tireless analyst who is always available: it can answer a quick question in seconds, run a multi-week project alongside you, and remember your preferences and history, so you don’t have to recontextualize your work at every turn.
With SuperAnalyst, you can:
- Work across AlphaSense: Search trusted content, analyze financial data, work with financial models, and use other AlphaSense capabilities.
- Remember your projects: Retain your project history, instructions, analytical approach, and prior work across sessions so you can continue where you left off.
- Create and update deliverables: Build reports, presentations, spreadsheets, financial models, dashboards, watchlists, and research memos grounded in source-linked intelligence. Ask SuperAnalyst to update deliverables with new data, and it automatically hands you current and polished outputs.
- Automate repeatable work with Skills: Use built-in Skills for workflows such as earnings analysis, due diligence, and competitive benchmarking. You can also create custom Skills based on your team’s processes and templates.
- Keep you in control: Review the sources and reasoning behind the work, validate key findings, and decide how the output is used.
How SuperAnalyst Compares to Generative Search
| Generative Search | SuperAnalyst | |
|---|---|---|
| Core job | Answer a question from the content | Run the research workflow behind a decision or deliverable |
| Shape of output | A cited answer, in the answer surface | An answer, or a formatted, downloadable artifact: memo, model, comp table, slide deck, dashboard |
| Depth | Multi-step retrieval and synthesis | Multi-step: plan, retrieve across sources, reconcile, compute, structure |
| Tools | Search and create slides | Search, structured financials, screeners, models, charting, file builders, web search — and Generative Search itself |
| State | Question by question | Conversational memory within and across sessions |
| Best for | "What did management say about pricing?" | "Build the pricing benchmark vs. these four competitors and put it in a deck." |
Generative Search works on a prompt. SuperAnalyst runs the entire workflow.
How to Get Started
Open the chat and tell SuperAnalyst what you need.
For example: "I need to review Oracle's latest earnings, specifically interested in what the Street is saying."
SuperAnalyst inherently understands what you need, calls on the right Skills and sources, and returns a sourced output — which you can easily turn into a deliverable by asking SuperAnalyst to turn it into a report, memo, or slide deck. Click any citation to review the original material.
The Building Blocks: Task, Skill, Artifact
Everything in SuperAnalyst breaks down into three connected pieces:
Task — The instruction you give it: one assignment, described in plain language. Example: "Build a Q1 earnings preview for NKE."
Skill — A Task you've saved to reuse. Configure it once, then run it again on a new company without rewriting the instructions. Example: "Map the competitive landscape to show where the target company wins."
Artifact — What comes back: the finished output. A single Task can deliver more than one Artifact at a time — for example, a summary memo and a companion slide deck from the same request.
Because SuperAnalyst is conversational, not silent, it will ask clarifying questions along the way instead of guessing at what you meant — so the Artifact you get back reflects what you actually needed.
Skill vs. Workflow Agent — two different things
Both let you save something and run it again, but they live in different places and save different things:
| Skill | Workflow Agent | |
|---|---|---|
| Lives in | SuperAnalyst | Generative Search ("My Agents") |
| What's saved | A multi-step Task — a whole delegated workflow | A single prompt — one question, saved to repeat |
| Trigger |
On-demand only, today Coming soon: scheduled to run automatically and reactive to market events |
On-demand, or scheduled to run automatically |
| Output | A structured, often multi-artifact deliverable | A refreshed answer, in the Generative Search format, including slides |
| Built for | Recurring assignments — earnings prep, comp sets, diligence | Recurring questions — "has anything changed on X?" |
The quick tell: if what repeats is a question, that's a Workflow Agent. If what repeats is an assignment with a deliverable at the end, that's a Skill.
Navigating SuperAnalyst
Chat with Your Agent
The chat is where you and your agent work together. You'll see how your agent is reasoning through a task, what it's doing step by step, and how each claim is backed by a citation you can click through to the original source. If something looks off-track, you can jump in at any point to redirect it.
Use Skills to Automate Repeat Work
Skills are reusable workflows you can hand to your agent — think of them as a saved playbook for a task you do often, like a comps update, a KPI pull, or pre-earnings prep. AlphaSense provides a library of ready-to-use Skills, and you can build, customize, and reuse your own so your agent handles recurring work the same way, every time. Call on a specific Skill using "/", or simply ask SuperAnalyst to do something and it will identify which Skill the task requires.
A custom Skill can be built:
- From scratch, using natural language
- From an uploaded document or template
- Automatically, after SuperAnalyst completes a workflow
- Through the Skill Builder
- Customized from an existing pre-built Skill
Uploading Files
You can also upload your own files directly into a task thread (drag-and-drop, or the + button). The file becomes part of that task's working context alongside AlphaSense's own content — but stays scoped to that task; it isn't added to the shared content library. File uploads have a per-file size limit of 50 MB.
Get Real Deliverables, Not Just Answers
SuperAnalyst doesn't stop at a chat response, it can build the actual output you need: an investment brief, an earnings summary, a competitive intelligence report, a financial model, a client-ready presentation or slide deck, a watchlist, or a research memo.
To build slides, ask SuperAnalyst directly in the chat, for example, "Turn this research into a client-ready deck with an executive summary slide," or "Build me a pitch deck draft for [Company]." SuperAnalyst will structure the narrative, populate slides with cited data and charts, and hand back a polished first draft you can refine. You can upload an existing deck and ask SuperAnalyst to use it as the template and it will always build slides that follow your guidelines.
These deliverables are saved and stay accessible, so you can come back, reuse, or edit them later. Every claim in a deliverable is traceable back to its source, so you can verify it first.
SuperAnalyst now draws on AlphaSense's unified Life Sciences data foundation alongside AlphaSense's premium market content. This means life sciences and healthcare teams can ask SuperAnalyst complex, multi-step questions like comparing competitor pipeline activity in a therapeutic area, and get synthesized, source-linked answers without leaving the platform or toggling to separate clinical databases.
Example Tasks & Chats
Not sure where to start? Here are a few examples of what you can ask SuperAnalyst to do:
Quick questions
- "What did the Street say about Oracle's last print?"
- "Summarize the key takeaways from [Company]'s most recent earnings call."
Research & synthesis
- "Draft a risk assessment for [company], include the latest 10k risk factors and cross-reference them against the last 3 peer earnings calls."
- "Compare how our top three comps have talked about margin pressure this quarter."
Building deliverables
- "Build me an earnings preview for [company]’s next print, I want to know what is likely to have changed from their previous quarter’s earnings call.."
- "Turn this research into a client-ready deck with an executive summary slide."
Staying on top of a project
- "Remind me what we concluded on the deal thesis last week."
- "Update me on whether a new peer transaction shifted our comps outside the current valuation range."