What it is

Letter AI is an AI-native revenue enablement platform: AI roleplay and certification paths, generated training and sales content, a company-indexed assistant (the Letter AI Agent), RFP automation and an AI Sales Room, all in one product. It was founded in 2023, is based in Chicago, and is run by co-founder and CEO Ali Akhtar. It raised a $10.6 million Series A led by Stage 2 Capital in October 2025, which put HubSpot’s founding CRO Mark Roberge on the board. Four months later, on 2026-02-24, it raised a $40 million Series B led by Battery Ventures, with Y Combinator, Lightbank, Northwestern Mutual Future Ventures, Stage 2 Capital and Formus following on. That brings the total to about $51 million. Named customers include Lenovo, Adobe, Novo Nordisk, Plaid, Zip, Kong and SolarWinds.

Letter Compass launched with the Series B. It records and transcribes calls on Zoom, Teams, Meet, RingCentral and GoTo without a separate recorder, scores them, tracks buyer sentiment and the buying committee per deal, and pushes deal-specific coaching, messaging and content to the rep. It reads deal context from the CRM, can run alongside Gong, and posts into Slack or Teams.

The question this page answers: now that Highspot is part of Seismic, is Letter AI a credible lighter alternative for teams that want enablement without a multi-year platform migration?

Why it shows up in RevOps stacks

The incumbents just merged. Seismic closed its merger with Highspot on 2026-08-18, and Highspot now ships as “Highspot by Seismic.” Teams up for renewal on either product are re-evaluating for the first time in years, and Letter AI is the newest vendor on most of those shortlists.
Content and training are generated, not just stored. The legacy platforms started as libraries you fill. Letter AI builds roleplay scenarios, training modules and first-draft sales assets from your own material in minutes. For a small enablement team, that changes what one person can ship in a quarter.
It moves enablement into the deal. Compass ties coaching and content to a specific opportunity and stage rather than a course catalog. That is the gap RevOps leaders usually point to when sales enablement completion rates look fine and win rates don’t move.
Reviewers rate it highly, on a small base. G2 shows 4.9/5 across roughly 75 reviews. Letter AI says it earned the highest Momentum Score of any enablement platform in G2’s Winter 2026 reports. Treat that as an early signal, not a track record.

Pricing

No public price list. Every deal is an annual quote, and the HubSpot app requires the Full Platform Subscription.

Vendr’s page for Letter AI, checked today, shows a median buyer paying $150,080 a year, with a range of $15,000 to $448,848. Vendr does not publish how many purchases sit behind that median, so read it as enterprise-weighted and directional. One G2 reviewer, a revenue enablement manager, described it as enterprise tooling at SMB prices, which fits the bottom of that range rather than the median.

Read that as: a team of 50 to 150 reps buying training, content and the Agent should anchor near the $15,000 floor of that range, not the median. Compass and Deal Pursuit rolled out across a few hundred sellers is what a median-sized contract looks like.

Best for

Enablement and RevOps leaders at B2B companies with roughly 30 to 500 quota-carrying reps, where a one- or two-person enablement team needs to produce onboarding, certifications and refreshed content faster than it can by hand. It is the strongest fit when you are on HubSpot, renewing Highspot or Seismic in the next 12 months, and want roleplay, content and deal coaching from one vendor without a six-month implementation.

Not for you if you need governed content distribution across many business units, regions and regulated brands. That is still Seismic’s job. Also not for you if your only gap is practice reps: a roleplay-only tool is cheaper and faster.

Versus the alternatives

Seismic (including Highspot by Seismic): the top two by share, now one company with 2,500 customers. Pick Seismic when content governance, brand control and analytics across a large, multi-division sales org are the requirement. Pick Letter AI when speed of content and training creation matters more than governance, and you would rather not sign a multi-year term during a product consolidation. See Seismic.
Mindtickle: the largest independent readiness platform. Pick Mindtickle when you need enterprise depth on certification and a single readiness score across thousands of reps, plus a vendor with a decade of deployments. Pick Letter AI when you want generated content and deal-level guidance in one lighter product. See Mindtickle.
Hyperbound: the fastest-growing entrant on the practice side, AI roleplay only. Pick it when practice is the entire problem and you want it running in weeks. See Hyperbound.
Gong already in place: do not pay twice for call recording. Run Compass alongside Gong, or scope Compass out of the order form and buy Letter AI for training and content only.

If none of these fit, the gap is usually ownership, not software. Someone still has to write the talk track, the battlecards and the certification rubric.

Watch-outs

Integration depth is uneven. HubSpot has a listed marketplace app with Compass dashboards over pipeline, deals and quota. ooligo found no public Salesforce AppExchange listing or setup documentation as of 2026-10-22, and G2 reviewers flag integration setup and partial Google Drive support. Guard: if you run Salesforce, have Letter AI connect it in a trial tenant and show Compass on your real opportunities before you sign, and write the CRM objects it must read into the order form.
A young vendor on a fast funding cadence. Two rounds in four months fund growth, but the roadmap and the team are still settling. Guard: keep the first term at 12 months, ask for a data-export clause covering content, training modules and call transcripts, and cap the renewal uplift in writing.
Generated content can be confidently wrong. AI-built training and assets reflect whatever source material you feed them. Guard: require an enablement or product-marketing review step before any generated module is assigned or any asset is shared externally, and audit a sample each quarter.
Compass records customer calls. Guard: confirm consent flows for every country you sell into, sign the DPA with explicit retention periods for recordings and transcripts, and keep AI call scores out of compensation decisions until HR and legal have reviewed them.

Edit this page on GitHub