Our verdict
HatchWorks AI suits a mid-sized US company that wants its data pulled together and an AI assistant or model built on top by one nearshore team at moderate hourly rates. Its case studies show systems in production for a franchise network, a home renovation investor and an IoT company, but many of its client reviews cover staff augmentation and recruiting, so buyers should confirm that the people assigned have built AI systems before.
The right buyer
- Companies whose data sits in separate systems and that want it unified before an assistant goes on top. For K9 Resorts it built a Databricks lakehouse across more than 52 franchise locations, a performance dashboard and K9 Chat, an assistant that answers plain-English questions from the same validated figures (K9 Resorts case study).
- Teams that want a machine learning model inside software their staff already use. Its three underwriting models for Grandview Homes run in Salesforce and cut after-repair value prediction error by 31% and time-to-sell error by 69% against the previous benchmark (Grandview Homes case study).
- Product companies adding a natural-language assistant to their own software. Cox2M’s director of data reported that the Kayo fleet assistant answered user questions with over 90% accuracy and arrived on time and within budget.
- Buyers who want US time zones at nearshore rates: its directory listing shows $50 to $99 an hour and a $25,000 minimum, and its engineers work from Costa Rica, Colombia, Peru and Brazil (About page).
Reasons to look elsewhere
- Many of its 29 client reviews cover staff augmentation and recruiting, so its 4.9 rating mixes that work with its AI projects.
- Its generative AI record is short. The company refocused on generative AI after November 2022 (FAQ), and the main result in its aviation case study, a 30% drop in recurring actuator issues, is a pilot example as stated by the client (Aero Star case study).
- The client review summary flags documentation as the area clients most want improved, which matters to a buyer planning to run the system in-house.
- Several of the larger engagements on its case studies page, such as a national commercial real estate firm and a mid-cap manufacturer, are anonymous, so the clients it identifies are mostly mid-sized companies.
HatchWorks AI, At A High Level
HatchWorks AI is an AI and data consulting firm in Atlanta, Georgia, that builds data platforms, AI assistants and machine learning models, mostly for mid-sized US companies.
- Founded in 2016 by Brandon Powell, who had spent 16 years in technology consulting and leadership, including at AT&T (Voyage ATL; About page).
- Rebuilt itself around generative AI after ChatGPT’s release in November 2022, retraining its teams and refocusing its services (FAQ). In 2024 it launched Generative-Driven Development, its method for building software with AI tools (Generative-Driven Development).
- Has US offices in Atlanta, Chicago and Dallas and Latin American offices in San José, Costa Rica, Bogotá and Medellín, Colombia, Lima, Peru and São Paulo, Brazil. It expanded into Costa Rica in 2019 and opened its Colombian offices in 2023 (About page; FAQ).
- Its partners include Anthropic, Google, Databricks and OpenAI. It joined Anthropic's Claude Partner Network in June 2026, embedding Anthropic-certified Forward Deployed Engineers to put Claude into production for enterprises (HatchWorks AI).
- It holds a 4.9 rating from 29 client reviews, and its directory listing splits its work into AI consulting (40%), AI development (40%), custom software (10%) and generative AI (10%); it also provides staff augmentation and recruiting.
Its documented client work includes:
- A Databricks lakehouse for K9 Resorts, a private-equity-backed dog boarding franchise with more than 52 US locations, joining operational data from Gingr with financial data from ProfitKeeper. On top it built the K9 Fetch dashboard and K9 Chat, an assistant on Anthropic’s Claude Sonnet and Databricks Genie; a review of more than 70 KPI question types found 61% answerable from the platform (case study).
- Three machine learning models for Grandview Homes, a Chicago company that buys, renovates and resells houses, predicting after-repair value, rehab cost and time to sell. The models run on Azure Machine Learning and show inside Salesforce, with monthly retraining (case study).
- The Kayo AI Assistant for Cox2M, an Atlanta IoT company, which answers fleet managers’ questions about trips, braking and mileage using retrieval-augmented generation on Google Cloud and Vertex AI (case study).
- AVA, an assistant for Aero Star Aviation’s aircraft technicians that answers from past work orders with citations and pulls exact values with text-to-SQL, built on Google Cloud (case study).
- The Meerir MVP for DronePort Network, which lets users query air traffic data in plain English, built between June and September 2025.
Scores
| Criterion | Weight | Score | Basis |
|---|---|---|---|
| Production track record | 20% | 7.5 | Systems in production for K9 Resorts, Grandview Homes and Cox2M; aviation and drone assistants at pilot or MVP stage. |
| Technical depth | 15% | 7.5 | Lakehouse data engineering, retrieval-augmented assistants, text-to-SQL and supervised models with retraining pipelines. |
| Industry depth | 10% | 6.0 | Work spread across franchising, real estate, aviation, IoT, finance and staffing, with no single deep vertical. |
| Delivery model and handoff | 15% | 7.0 | US leads with nearshore teams in matching time zones; client reviewers praise project management and flag documentation as the weak spot. |
| Commercial clarity | 10% | 6.5 | Hourly range and minimum project published in its directory listing; no prices on its own site. |
| Platform independence | 10% | 7.5 | Claude Partner Network member since June 2026; also partners with Google, Databricks and OpenAI; client builds span Databricks, Google Cloud and Azure. |
| Stability and scale | 10% | 7.0 | Operating since 2016 with offices in four countries; Inc. 5000 company in 2021. |
| Independent evidence | 10% | 7.0 | 29 client reviews averaging 4.9, many of them for staffing rather than AI work; client quotes with full attribution in case studies. |
| Overall | 100% | 7.1 | Weighted average |
Where HatchWorks AI fits
| Need | Fit | Why |
|---|---|---|
| A unified data platform with an AI assistant on top | Strong | K9 Resorts lakehouse, dashboard and K9 Chat across more than 52 locations. |
| Natural-language assistants inside a company’s own product | Strong | Cox2M Kayo fleet assistant; DronePort Meerir MVP. |
| Predictive models inside existing business software | Good | Three Grandview Homes underwriting models running in Salesforce. |
| Extra engineers for an in-house team | Good | Staff augmentation and recruiting account for many of its client reviews. |
What clients and leaders say
“For the first time, we have a single, trusted view of performance across every K9 Resorts location. That changes how we make decisions as a brand.”
“Offer pricing is where our margin lives. Having model-driven predictions inside Salesforce gives our underwriters a sharper, more consistent starting point than the benchmark we relied on before.”
“HatchWorks AI's Gen AI Innovation Workshop has transformed how we think about Gen AI by getting our entire team on the same page and speaking the same language.”
“Putting Claude into production is where the real value shows up for our customers. Enterprises do not have an AI access problem. They have an AI adoption problem.”
Who leads HatchWorks AI
Sources (11)
- HatchWorks AI: homepage
- HatchWorks AI: About us
- HatchWorks AI: FAQ
- HatchWorks AI: Case studies
- HatchWorks AI: K9 Resorts case study
- HatchWorks AI: Grandview Homes underwriting case study
- HatchWorks AI: Cox2M Kayo AI Assistant case study
- HatchWorks AI: Aero Star Aviation AVA case study
- HatchWorks AI: Generative-Driven Development
- Voyage ATL: Meet Brandon Powell of HatchWorks (July 25, 2018)
- HatchWorks AI: joins the Claude Partner Network (June 23, 2026)