Volatility has increased across technology markets as the AI-led trade fragments into winners and losers. Capital has rotated out of some overheated hardware plays and back into enterprise-focused software, advisory services, and cybersecurity, producing abrupt reversals and pocketed rallies rather than steady sectorwide advances.
Against this backdrop, a rules-driven portfolio selection system that combines machine learning with institution-grade valuation inputs has been identifying companies it considers materially undervalued. Subscribers to the platform's premium tier, who gained access during a limited-time Flash Sale at a rate below $8 per month, are positioned in a growing set of those high-conviction names.
Recent market action illustrated the shift. Gartner Inc (NYSE: IT) rallied 9.74% in a sharp software-orientated relief move, and Fortinet Inc (NASDAQ: FTNT) added 9.04% as institutional capital rotated back into advisory and cybersecurity franchises while corporate buyers reassess AI spending plans. The system's recent selections also outperformed: Everforth Inc (NYSE: EFOR) jumped another 8.20%, extending its cumulative gain to 93.91% since it was first identified by the models in July.
How the selection engine operates
The selection framework - branded as ProPicks AI - evaluates thousands of global equities each month by processing more than 150 fundamental inputs and a broad set of quantitative models. It incorporates aggregated Fair Value estimates derived from 17 institutional-grade valuation approaches to help surface securities trading at large discounts to intrinsic estimates.
That combined approach aims to find the companies where a gap between current price and underlying fundamentals suggests a material medium-term re-rating opportunity. Premium subscribers gain access to an AI-curated list of up to 20 high-conviction stock picks per strategy on a monthly cadence, with the system rebalancing each month to add new candidates, retain outperformers, and remove positions that no longer meet the criteria.
Highlighted undervalued opportunities
The models returned a list of several deeply discounted setups. Among them:
- Everforth (NYSE: EFOR) - flagged with 51.9% Fair Value upside and discussed in detail below.
- [LOCKED TRAVEL GIANT] - 50.6% Fair Value upside (full details available to members).
- [LOCKED FINANCE LEADER] - 46.4% Fair Value upside (full details available to members).
In addition to these examples, the system noted several strong performers through September, including large month-to-date moves and extended gains since selection:
- Inspire Medical Systems (NYSE: INSP): +20.81% in September alone.
- Alphatecs (NASDAQGS: ATEC): +17.92% in September alone, and +22.57% since first picked.
- CVR Energy (NYSE: CVI): +13.44% in September alone, and +36.96% since first picked.
- Everforth (NYSE: EFOR): +12.45% in September alone, and +93.91% since first picked.
- HP Inc (NYSE: HPQ): +9.67% in September alone, and +24.75% since first picked.
Everforth - a low-valuation setup backed by contract wins and cash generation
Everforth was identified by ProPicks AI in July at a valuation level the models classified as deeply disconnected from the company's underlying performance. The selection preceded a notably strong Q2 report. On a top-line basis the company produced $1.007 billion in revenue for the quarter. Operational metrics were positive: the platform generated $96.7 million in adjusted EBITDA and $46.3 million in free cash flow for the period.
Management has used that cash flow to support capital returns. The company is operating under a $1 billion buyback program and completed a $77.5 million tranche that repurchased roughly 1.79 million shares. Leadership also guided toward potential Q3 revenue of up to $1.024 billion and projected EBITDA margins near 10.3%.
Commercial momentum included double-digit growth in consulting bookings as corporate clients accelerated enterprise software rollouts. The firm's defense business anchored additional resilience, with the company securing a $115 million U.S. Army AI contract that the models pointed to as evidence of an embedded public-sector AI capability and a sticky revenue stream.
Using its aggregated valuation inputs, the platform assigns Everforth a Fair Value of $51.71 - implying roughly 51.9% upside from the level used in the analysis.
What the models detected prior to the rally
The selection engine combines investor-grade financial research with machine-learning assessments across a global universe of more than 60,000 equities. For Everforth, the signals that triggered the buy included:
- Deep valuation disconnect - At the time of selection the stock was trading near $18, down roughly 63% year-to-date and at about 29% of its 52-week high. The company appeared inexpensive on multiple measures: an adjusted price-to-earnings near 6x and a market price at less than half of reported book value. With management projecting $5.74 in EPS by 2027, the shares were trading at roughly 3x forward earnings.
- Strong cash generation and active capital returns - The core business produced $1.007 billion in quarterly revenue with $96.7 million in adjusted EBITDA and $46.3 million in free cash flow, accompanied by a completed $77.5 million buyback tranche removing 1.79 million shares from the float.
- High-value enterprise and AI positioning - Double-digit consulting booking growth from enterprise software deployments, together with a durable defense backlog including a $115 million U.S. Army AI research contract, suggested a sticky revenue mix and potential margin expansion driven by higher-value digital engineering work purchased through the Quinnox acquisition.
- Upside catalysts - A range of external price targets and independent fair-value models cited mid-to-high double-digit upside, with referenced Wall Street targets near $29 and other fair-value estimates around $31 that the models interpreted as supporting 65% to 70% potential upside in some scenarios. Additional catalysts included the company's inclusion in the Russell 2000 index and notable insider share purchases.
Performance track record and membership access
Since the formal launch of the AI models in November 2023, the system's published strategies have compounded into cumulative gains recorded at +192.92%, outpacing the S&P 500 by +113.03% over the same period, according to the system's performance tracking. That historical record is presented as a realized figure since the models began operating.
For those interested in the full list of September selections and the model-level rationales, premium members have direct access to the complete dossier. The platform has periodically offered subscription discounts, including a Flash Sale that reduced membership costs by as much as 55% for the promotion's duration.
Subscription access routes were listed for both app and web users in the original presentation for prospective members who wished to view the full stock lists and model outputs.
Methodology in brief
Each month the AI engine scans thousands of global equities, processing more than 15 years of historical financial data through over 150 quantitative models. It ranks names by a combination of valuation signals, momentum, and forward-looking growth metrics to form strategies that are equally weighted across all selected stocks. That equal-weighting is intended to provide a transparent benchmark for tracking the model's overall performance while allowing individual users to adjust exposures as they see fit.
Key takeaways
- The AI-driven stock selection system has surfaced multiple deeply discounted names as the market rotates into enterprise software and cybersecurity.
- Everforth, a recent top pick, rallied roughly 93.91% since selection and remains flagged with approximately 51.9% Fair Value upside by the model's aggregated estimates.
- Several other selections have produced double-digit monthly gains, highlighting the system's focus on valuation disconnects and cash-generative business models.
Note on subscription pricing and model performance - Subscription prices referenced in the platform's materials are stated as accurate at the time of publication and may vary by region. Performance figures cited reflect realized returns since the official model launch in November 2023.