Northvalley IntelligenceSoftware for everyone.

Marietta, Georgia

Northvalley Intelligence for government buyers.

Northvalley Intelligence builds software for organizations that need it to work with the systems they already run: web and mobile applications with custom integrations, agentic applications, and data science, machine learning and data cleansing.

Download the capability statement (PDF)

Core competencies

Software development

Web and mobile applications plus custom integrations with the systems a buyer already runs.

Agentic applications

Software agents that take actions inside a business's own systems, with voice and chat front ends and agent-callable backends (MCP).

Data science, machine learning and data cleansing

Data cleansing, entity normalization and matching, ontology, ranking and search relevance, and machine learning models, on messy real-world data.

How an engagement runs

  1. 1Conversation with the founder.
  2. 2Written scope and acceptance criteria.
  3. 3Build in short increments on a staging environment.
  4. 4Release through CI, with documentation and handover.

Differentiators

  • Entity recognition and natural-language processing on messy real-world data: normalization and matching systems at CareerBuilder, a product ontology and machine learning ranking at The Home Depot, and catalog search and discovery with generative AI at Genuine Parts Company.
  • Data cleansing as a repeatable method: profile, normalize, match and measure, with an evaluation set agreed before the cleanup starts.
  • Agent-callable delivery: systems built so a customer can reach them from the assistants they already use. See the live demo under Capabilities.
  • Enterprise release discipline at small-business scale: a staging environment, scenario tests and CI-gated deploys on every engagement, with the founder (Ph.D., 22 years in software) doing the hands-on work.
  • Models can be trained and run on local hardware, so data stays in-house where a client needs it.

Company data

Legal name
Northvalley Intelligence LLC
Entity
Georgia limited liability company, founded June 2026
Location
Marietta, Georgia
NAICS (primary)
541511 — Custom Computer Programming Services
NAICS (secondary)
541512 — Computer Systems Design Services
PSC
DA01 — IT and Telecom - Business Application/Application Development Support Services (Labor); DH01 — IT and Telecom - Platform Support Services: Database, Mainframe, Middleware (Labor); DD01 — IT and Telecom - Service Delivery Support Services: ITSM, Operations Center, Project/PM (Labor)
Contracting role
Prime contractor or subcontractor
Unique Entity ID (UEI)
JX76HA57NCJ8
CAGE code
26B16
SAM.gov registration
Active
Download the capability statement (PDF)

Two pages, mirrors this page. Updated when the page changes.

Past performance

Canon Insurance Advisers

  • Client: Canon Insurance Advisers, LLC (Georgia insurance agency).
  • Role: direct engagement, founder as sole developer.
  • Period: July 2026 to present (currently active).
  • Situation: a Georgia insurance agency quoted each home and auto prospect by hand, re-keying the same details for several hours per quote.
  • Task: cut quoting time without replacing the agency's existing tools.
  • Action: a one-page intake form feeds an automated quoting run that completes the rating steps unattended, with guard rails that never bind, purchase or order a report; an earlier phase added a retention board that flags policies at risk of lapsing before they end.
  • Result: several hours of manual quoting became a few seconds of form filling and a few minutes of automated quote generation.
  • Reference: available on request.

Smart Cash Sheet

  • Client: Smart Cash Sheet.
  • Role: ongoing engineering.
  • Period: September 2026 to present (currently active).
  • Situation: a personal cash-flow app on the web with iOS and Android apps, run by a solo founder, shipped about once a month, each release carrying weeks of accumulated change and risk.
  • Task: take over the engineering and ship faster without breaking production.
  • Action: a staging environment, scenario tests and a three-phase release process; bank-data import, subscriptions and billing, and the safety rules around the in-app money coach.
  • Result: the monthly cycle became two production releases in one week, turning the founder's roadmap into shipped features faster.
  • Reference: available on request.

Public write-ups: Client Work

Key personnel

Ferosh Jacob, Ph.D. Founder and technical lead

Northvalley Intelligence's work is performed by its founder. Before founding the company in June 2026 he spent 22 years in software, moving from software engineer to senior engineering manager leading cross-functional teams of more than 40 engineers, managers, data scientists and machine learning engineers, across the retail, healthcare, recruiting and automotive-parts domains.

Senior Engineering Manager

July 2024 – June 2026

Genuine Parts Company, Atlanta, GA

Led a cross-functional domain of four squads on catalog, search and discovery, using generative AI and classical machine learning. Built a first-of-its-kind agent-based auto-parts assistant on an Agent Development Kit, working with UX and platform teams through A/B testing and production rollout.

Senior Engineering Manager

January 2021 – June 2024

Kaiser Permanente, Atlanta, GA

Led a cross-functional team of managers, data scientists, machine learning engineers and software engineers responsible for search relevance and platform performance for internal and external clients. Directed the migration of KP.org search from a vendor product to a homegrown open-source solution, transitioning the endpoints with minimal disruption to other services. Two refereed papers followed on automated search-relevance evaluation (AMIA 2022; CIKM 2022).

Staff Software Engineer

September 2015 – January 2021

The Home Depot, Smyrna, GA

Led data-driven initiatives in search relevance, product ranking and ontology development for the online store: machine learning ranking models, a search-evaluation framework that measured feature impact before deployment, and a product ontology that improved search precision and recall from customer behaviour and external datasets.

Senior Software Engineer

September 2013 – September 2015

CareerBuilder, Norcross, GA

Led a machine learning school and company normalization system that identified and standardized educational institutions, companies and other entities from unstructured text, and developed profile merging and matching across multiple data sources for the Recruitment Edge project. Co-author of the 2016 Springer chapter "Data Science and Big Data Analytics at CareerBuilder".

Software Engineer

June 2004 – July 2007

Stabilix Private Solutions Ltd., Trivandrum, India

First professional role, supporting a United States insurance client from India and building the security module of Health Companion, a generic application and claim-processing system for insurance companies.

Education and research
Ph.D. in computer science, The University of Alabama (2009–2013) — thesis "Modeling Computation-Intensive Problems for Heterogeneous Computing and Improved Source Code Maintenance", advisor Dr. Jeff Gray; the research was supported by an Air Force Research Laboratory STTR grant. M.S. in computer science, Clarkson University (2007–2009) — thesis "CSeR: A Code Editor for Tracking and Visualizing Detailed Clone Differences". B.Tech. in electronics and communication, National Institute of Technology Calicut, India (2000–2004). Ph.D. intern at Pacific Northwest National Laboratory (2012): two domain-specific languages that let scientists describe and deploy signature-discovery workflows, integrated into the laboratory's pipeline and published in IEEE Computing in Science & Engineering (2014).
Teaching
Part-time Assistant Professor of computer science, Kennesaw State University, since May 2017 — information retrieval, big-data analytics, text mining, data warehousing and mining, requirements engineering, software testing, software project management and algorithm analysis.
Publication record
Twenty-seven published items since 2009: five refereed journal articles, a Springer book chapter, sixteen refereed conference papers, four refereed workshop papers and a conference tutorial — in IEEE Computing in Science & Engineering, The Journal of Supercomputing, Computer Physics Communications, the Journal of Parallel and Distributed Computing, IEEE Big Data, the AAAI industry track, AMIA and CIKM. Full list at feroshjacob.github.io.