DyscernAI is building an agentic AI layer that ingests unstructured market content and evaluates it the way a company's own analyst would — trained on the customer's expertise, not off-the-shelf. The bet is that the next generation of market-intelligence tools is not a better Bloomberg, but a per-customer agent stack that replaces manual news triage and PR/competitive scanning inside enterprise sales, product, and executive teams.
The team is unusually strong for a Seed: Brandon Duncan (Harvey Mudd, prior operator experience), Tim Gernitis (Bucknell; ex-Tend Inc.), and Steve Yankovich (advisor-level, former Vice President at eBay running mobile / commerce). Company has been in SH's pipeline since June 2025 with monthly cadence. Raising a $4M SAFE with a second $3M planned for late 2026 / Q1 2027 that SH will have access to. Validation is still thin on the customer side — needs proof of paid enterprise deployment before the seed is fully de-risked.
Enterprise research-and-intelligence AI is a fast-growing subset of the broader legal / enterprise AI market — $5.6B in 2026, growing 22% CAGR to $12.5B by 2030 (Web). AlphaSense's public trajectory (acquired Sentieo in 2022, ~$4B valuation, rolled out an autonomous AI "Channel Checks" agent in 2025) confirms that the market is ready to pay for agentic research (Web). The bull case is that per-company customization creates room for a challenger below AlphaSense's price point.
| Player | Positioning | Funding / Stage | Edge vs. them |
|---|---|---|---|
| DyscernAI | Per-customer agentic evaluator over unstructured market content | Seed | — |
| AlphaSense | AI research + agentic Deep Research + Channel Checks | ~$4B valuation | AlphaSense is the entrenched incumbent; DyscernAI must be cheaper and easier to customize |
| Bloomberg / BQuant + AI | Terminal + LLM tools | Public / private | Bloomberg is the default; DyscernAI wins outside financial services in product / sales / competitive intel |
| Perplexity / OpenAI Deep Research | General-purpose research agents | Late-stage | Perplexity is horizontal; DyscernAI targets vertical evaluators tuned to specific enterprise workflows |
| Klue / Crayon | Competitive intelligence SaaS | Series B/C | Klue is human-curated; DyscernAI aims for agentic auto-evaluation |
Moat: The moat is the "evaluator engine" — customer-specific tuning that captures internal analyst expertise and gets stickier the longer it runs. That's a data-network moat if DyscernAI can capture the internal expertise; a distribution problem otherwise.
Comps: AlphaSense's Sentieo acquisition (2022), S&P Global acquiring Kensho for $550M (2018), and Reuters buying similar analytics tools. Salesforce / Microsoft / ServiceNow are all natural enterprise-AI acquirers as agentic workflows land inside CRM and productivity stacks. Bloomberg has historically preferred build over buy but has been more acquisitive in the AI era.
Team is progressing on evaluator engine and pipeline; SH continues to track ahead of a formal check decision.
Update on evaluator engine and enterprise trials.
Adding senior operator credibility with Steve Yankovich (ex-eBay VP) — first meaningful team signal.