Use case · Hedge funds

The end of manual data entry for your investment team.

OpenHelm closes the last mile of analysis with a human operator in the loop: we update your financial models from the filings, clean the messy alternative-data feeds you pay for, and turn 300 overnight emails into a thesis-aware briefing. The agent drafts, our operator verifies, you approve the merge.

Free first month on a defined brief, benchmark before you commit
Model: NVDA Q3
Alt-data feed
Morning brief
Filings monitor
Update the NVDA model from the 10-Q, same conventions as our master file.

now

Extracted actuals from NVDA 10-Q· 14 line items
Drafted cell updates in a sandbox copy
Operator verifying mappings vs master model
Diff waiting for your approve-merge
Where shall we steer?
Operator · live

A fraction

Of a £180–250k all-in hire

Week one

Time to first output

Operator-checked

Every model update verified before it merges

Already in production with

Camio logoSmoothie Wars logoSummited logoAcademic Edge Tuition logoKicks Therapy logoCamio logoSmoothie Wars logoSummited logoAcademic Edge Tuition logoKicks Therapy logo

What we do

The three jobs we lead with, and the ones we cover alongside.

Model maintenance

Actuals extracted from the 10-Q/8-K, cell updates drafted in a sandbox copy, mappings operator-verified, delivered as a diff you approve-merge. We never silently overwrite your formulas.

Alternative-data triage

Messy credit-card, web-scrape or pricing feeds normalised, mapped to tickers, missing values handled, output as a clean time-series ready for Snowflake or Excel.

Thesis-aware morning briefing

Overnight filings and news filtered through your written thesis per ticker. You only hear about what supports or challenges the position, not every Form 4.

Filings & risk-factor monitor

Material risk-factor changes and accounting divergences across your coverage, flagged with the page reference.

Earnings triage (alongside)

Guidance changes and key quotes surfaced. Useful, but not the wedge, the giants already do this well.

Comp set refresh (alongside)

Peer multiples kept current. A commodity feature we include, not something we ask you to pay for.

What you receive

What it actually looks like.

10-Q actuals mapped to your master model. You review the diff, then merge.

Model update, awaiting approve-merge

Live

Line itemModelFrom 10-QStatus
Data Center rev30,77130,818Verified
Gross margin %74.6%74.6%Verified
Opex4,2874,293Verified
Q4 guide (mid)37,00037,500Flag: review

Posted when an alt-data feed lands and is cleaned.

#

alt-data

OH

OpenHelmnow

Pricing feed ingested and ready

  • Ingested 14,000 rows of scraped pricing data.
  • Cleaned, deduped, missing values handled.
  • Mapped to 4 tickers: TSLA, F, GM, RIVN.
  • Clean time-series written to Drive · ready for Snowflake.

Filtered through your thesis. Signal, not noise.

Inbox

2 new

OH

OpenHelm

Thesis

NVDA guided up: challenges your datacenter cap-ex bear case

Mgmt raised Q4 datacenter guide 2%. Directly contradicts the slowdown thesis on file.

OH

OpenHelm

Muted

MSFT 8-K, no thesis impact, logged

Routine material-agreement disclosure, p.4. Filed, not flagged for action.

The same analytical output at a fraction of the headcount cost.

Research analyst
OpenHelm
Annual cost
£180,000–£250,000 all-in
Custom quote
Time to first output
4–6 months
Week one
Coverage ramp
6+ months
Briefed on your universe from day one
Attrition risk
High (18–24 month average tenure)
None
Management overhead
Significant (training, feedback, career management)
Minimal
Scalability
Add another headcount
Adjust scope of service

Pricing is bespoke and quoted to scope, but the comparison is not £180k+ all-in for a junior hire, plus ramp and recruiting fees, against a managed service that delivers from week one.

What clients say

OpenHelm basically gave us an entire marketing department overnight. This is what it feels like to punch above your weight!
Dr Thom Van Every

Dr Thom Van Every

Founder · Smoothie Wars

Pricing

Two ways to bring OpenHelm to the research analyst workload.

Self-serve, where your team drives OpenHelm directly, or a managed service, where a dedicated operator runs the brief on your behalf and stays in close contact across calls, email, and Slack.

Self-serve

Every tier includes every feature and unlimited seats.

Basic

For small teams getting started with AI at work.

  • Every feature, no gates
  • Unlimited seats
  • Voice + text interface
  • Isolated credential vault
Book a demo
Most popular

Pro

For teams putting OpenHelm to daily work across a few functions.

  • Priority onboarding
  • Team-shared memory & history
  • Read/write across shared Drives
  • Slack + MS Teams integration
Book a demo

Max

For teams running OpenHelm continuously across the business.

  • Priority support
  • Higher per-job concurrency
  • Custom workflow templates
  • Quarterly review with the team
Book a demo

Managed Service

A dedicated operator on your account, calibrated to the research analyst brief.

For teams that want a real human in the loop. Hop on calls, share data over email or Slack, and have outputs reviewed before they reach you. Pricing is bespoke to your scope and quoted on enquiry.

  • A dedicated human operator on your account, available by call, email, or Slack
  • Calibrated to your brief, format, and tone
  • Outputs reviewed by a human before delivery
  • First month free without commitment, then a custom quote

Custom quote

Quoted to your brief

12-month contract, first month free without commitment.

Enquire for a quote

Twenty minutes on a call, real scoping, then a quote.

What the output actually looks like.

Sample

Model update diff

A split view: the 10-Q excerpt on one side, the exact cells proposed for update on the other, each operator-verified. You click approve-merge; the formulas in your master model are never touched without sign-off.

Sample

Clean alt-data time-series

A messy scraped feed turned into a strict schema: entity names mapped to tickers, missing values handled, deduped, and written as a clean CSV ready for Snowflake or your model, with a pipeline-health note.

Sample

Thesis-aware briefing

Not "NVDA guided up" but "NVDA guided up, and it challenges your datacenter cap-ex bear case." Overnight news scored against the thesis you hold per ticker, so you read signal, not 300 emails.

Sample

Filings & risk-factor alert

A material risk-factor change or accounting divergence in a coverage name, flagged with the page reference and one line of context.

Frequently asked questions

Built for funds where output quality is non-negotiable.

The operator model: every output is reviewed by a human with role-relevant experience before delivery. This isn’t a chatbot.

The audit trail: every action logged, every output timestamped. If compliance asks what generated a research note, there’s a clear answer.

Data security: credentials stored in isolated vaults with scoped permissions. Nothing leaks sideways between clients.

The trial structure: first month at no charge on a defined brief, so you can benchmark quality against your own standards before committing.

Before you commit to the headcount, run one month of output.

We’ll work through a defined brief from your actual coverage universe for the first month at no cost. You benchmark the output against your standards. If it meets them, you enter a 12-month arrangement at a fixed monthly fee. If it doesn’t, you’ve lost nothing.