General Agency

What do they actually do

General Agency (now operating as Altrina) sells Tessa, a browser‑native AI coworker available as a web app and API. You describe a task in plain English and Tessa generates an editable workflow graph, then executes the steps in a real, logged‑in browser session. It can authenticate to sites, reuse a browser profile, navigate dashboards, and extract structured data like invoices and large tables. Runs can be started on demand, scheduled, or triggered via API, and the system keeps a full action log for review and audit (heytessa.ai, altrina.com).

Typical use: enter a goal (e.g., “send me a daily stock report” or “extract invoices”), review or tweak the generated steps, and let Tessa run while you monitor or take over if needed. The product ships workflow templates (invoice extraction, stock monitoring, support‑ticket flows) and exposes an endpoint for running its browser agent (heytessa.ai, altrina.com, YC Launch).

They publicly advertise strong scores on browser‑agent benchmarks and emphasize enterprise controls like multiple authentication options and full audit trails. They also list “teach by screen recording/video” and Google Workspace (Gmail/Calendar/Drive) integrations as coming soon (YC company page, altrina.com, heytessa.ai).

Who are their target customer(s)

• Finance / accounts‑payable teams: Hours lost logging into vendor portals, downloading invoices, and reconciling line items; manual clicks and re‑entry cause delays and errors. Tessa’s logged‑in browser automation and invoice/table extraction reduce this work (heytessa.ai, altrina.com).
• Operations / business‑ops teams: They run multi‑step processes across many web apps but lack engineering resources to build reliable automations, so work piles up or gets outsourced. Plain‑English workflows and editable graphs let non‑technical ops automate and inspect runs (heytessa.ai, altrina.com).
• Legal and compliance teams: They need accurate contract/table extraction and auditable trails for review; manual processes are slow and error‑prone. Tessa records actions and provides traceability to reduce risk (altrina.com).
• E‑commerce and marketplace managers: They must monitor prices/listings and pull reports from multiple seller dashboards behind logins; manual work and basic scrapers are brittle. Tessa authenticates, navigates dashboards, and extracts structured data on a schedule (heytessa.ai, altrina.com).
• Product teams building vertical apps: They want embedded, reliable browser automation without maintaining fragile scrapers. Tessa’s API exposes a hardened browser agent they can integrate into domain‑specific apps (YC company page, altrina.com).

How would they acquire their first 10, 50, and 100 customers

• First 10: Hand‑sell 4–6 week paid pilots to mid‑market finance/AP, ops, and e‑commerce teams using invoice‑extraction and logged‑in browser demos; include custom template setup, tight founder‑led onboarding, and a discounted trial for a testimonial and metrics on hours saved/error reduction (heytessa.ai, altrina.com, YC company page).
• First 50: Package early pilots into repeatable templates (AP, price monitoring, contract extraction), offer low‑friction self‑serve trials, and add 1–2 growth reps for targeted outbound and community channels. Run webinars and promote an API/partner track using pilot outcomes to drive signups (heytessa.ai, YC company page).
• First 100: Build connectors to accounting/ERP and CRM tools, recruit RPA consultancies/BPOs as resellers, and standardize subscription + usage pricing. Add an enterprise AE and customer success to convert mid‑market pilots into annual contracts using audit/compliance features and case studies as proof points (altrina.com, heytessa.ai).

What is the rough total addressable market

Top-down context:

Tessa sits across RPA, intelligent document processing, AP automation, workflow/BPA, and web‑automation. Public estimates put these at roughly: RPA ~$18B (2024), IDP ~$7.9B, AP automation ~$3–6B, workflow/BPA ~$15B+, and web scraping < $1B—together indicating a broad $40–50B universe with overlap (Fortune BI – RPA, Fortune BI – IDP, Grand View – AP, Persistence MR – BPA, Market.us – Web scraping, Tessa site).

Bottom-up calculation:

From those adjacent markets, restrict to use cases that require logged‑in, browser‑native automation, high‑quality document/table extraction behind paywalls, and an embeddable agent API; then discount overlapping categories by ~40–60%. This yields an estimated SAM of ~$15–25B. A reasonable early SOM is 0.1%–0.5% of SAM, or roughly $20M–$100M ARR over 3–5 years with strong execution and enterprise readiness (Tessa site, YC company page, market reports above).

Assumptions:

• Large overlap across RPA, IDP, BPA, and AP markets; apply a 40–60% discount to avoid double counting.
• Only a subset of automation buyers need logged‑in browser agents, auditability, and an embeddable API.
• Enterprise adoption depends on reliability, security, and compliance posture (e.g., SOC 2/GDPR/HIPAA).

Who are some of their notable competitors

• UiPath: Enterprise RPA platform for automating desktop and web processes with heavy IT governance. Overlaps on automation scope, but it targets IT‑led deployments vs. Tessa’s browser‑native, natural‑language agent for ops/API use cases (heytessa.ai).
• Apify: Cloud platform for running browser scrapers/actors at scale with an API and prebuilt jobs. Similar on logged‑in scraping and scheduling, but positioned as a developer tooling stack vs. Tessa’s NL workflows and editable graphs (altrina.com).
• PhantomBuster: No‑code web automation and data extraction, popular in lead gen and social. Strong for predefined connectors; Tessa aims at more decision‑heavy back‑office workflows with enterprise auditability (heytessa.ai).
• Diffbot: API that turns public web pages into structured data. Competes on extraction quality for public pages but doesn’t operate a logged‑in, interactive browser agent or multi‑step workflows like Tessa (altrina.com).
• DIY + open‑source (Playwright + agent frameworks): Teams can build custom browser automation with Playwright and open‑source agents (e.g., Auto‑GPT). Offers flexibility but requires engineering to harden logins, reliability, scheduling, and audit/enterprise features that Tessa provides out of the box (Auto‑GPT, heytessa.ai).