Category: Cost-Effective Enterprise AI

40-article series on cost-effective AI implementation in enterprise

Agent Cost Optimization as First-Class Architecture: Why Inference Economics Must Be Designed In, Not Bolted On

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18916800  39stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted40%○≥80% from verified, high-quality sources
[a]DOI20%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed30%○≥80% have metadata indexed
[l]Academic20%○≥80% from journals/conferences/preprints
[f]Free Access40%○≥80% are freely accessible
[r]References10 refs✓Minimum 10 references required
[w]Words [REQ]3,176✓Minimum 2,000 words for a full research article. Current: 3,176
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18916800
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (31 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

In 2026, inference costs account for 85% of enterprise AI budgets, yet most agentic system architectures treat cost optimization as an operational afterthought rather than a foundational design constraint. This paper argues that agent cost optimization must be elevated to a first-class architectural concern — embedded in system design decisions from the ground up alongside correctness, reliabil...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18916800 39stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted40%○≥80% from verified, high-quality sources
[a]DOI20%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed30%○≥80% have metadata indexed
[l]Academic20%○≥80% from journals/conferences/preprints
[f]Free Access40%○≥80% are freely accessible
[r]References10 refs✓Minimum 10 references required
[w]Words [REQ]3,176✓Minimum 2,000 words for a full research article. Current: 3,176
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18916800
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (31 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Open-Source vs Proprietary LLMs: Real Enterprise Economics

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18894954  34stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources7%○≥80% from editorially reviewed sources
[t]Trusted21%○≥80% from verified, high-quality sources
[a]DOI14%○≥80% have a Digital Object Identifier
[b]CrossRef7%○≥80% indexed in CrossRef
[i]Indexed29%○≥80% have metadata indexed
[l]Academic14%○≥80% from journals/conferences/preprints
[f]Free Access21%○≥80% are freely accessible
[r]References14 refs✓Minimum 10 references required
[w]Words [REQ]2,188✓Minimum 2,000 words for a full research article. Current: 2,188
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18894954
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]50%✗≥80% of references from 2025–2026. Current: 50%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (23 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

The choice between open-source and proprietary large language models (LLMs) is one of the most consequential economic decisions facing enterprise technology leaders in 2026. While proprietary APIs from OpenAI, Anthropic, and Google offer immediate access to frontier capability with zero infrastructure overhead, the true total cost of ownership (TCO) diverges sharply from sticker pricing at scal...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18894954 34stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources7%○≥80% from editorially reviewed sources
[t]Trusted21%○≥80% from verified, high-quality sources
[a]DOI14%○≥80% have a Digital Object Identifier
[b]CrossRef7%○≥80% indexed in CrossRef
[i]Indexed29%○≥80% have metadata indexed
[l]Academic14%○≥80% from journals/conferences/preprints
[f]Free Access21%○≥80% are freely accessible
[r]References14 refs✓Minimum 10 references required
[w]Words [REQ]2,188✓Minimum 2,000 words for a full research article. Current: 2,188
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18894954
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]50%✗≥80% of references from 2025–2026. Current: 50%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (23 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Bridging the Gap: Startup Workflows for AI Productivity Integration

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18868149  35stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted27%○≥80% from verified, high-quality sources
[a]DOI13%○≥80% have a Digital Object Identifier
[b]CrossRef7%○≥80% indexed in CrossRef
[i]Indexed33%○≥80% have metadata indexed
[l]Academic7%○≥80% from journals/conferences/preprints
[f]Free Access20%○≥80% are freely accessible
[r]References15 refs✓Minimum 10 references required
[w]Words [REQ]2,814✓Minimum 2,000 words for a full research article. Current: 2,814
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18868149
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]43%✗≥80% of references from 2025–2026. Current: 43%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (24 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

Startups occupy a paradoxical position in the 2026 AI landscape: unburdened by legacy infrastructure, yet resource-constrained in ways that make AI adoption both essential and precarious. Gartner projects that 40% of enterprise applications will incorporate task-specific AI agents by end of 2026, up from less than 5% in 2025 — a near order-of-magnitude leap that compresses traditional adoption ...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18868149 35stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted27%○≥80% from verified, high-quality sources
[a]DOI13%○≥80% have a Digital Object Identifier
[b]CrossRef7%○≥80% indexed in CrossRef
[i]Indexed33%○≥80% have metadata indexed
[l]Academic7%○≥80% from journals/conferences/preprints
[f]Free Access20%○≥80% are freely accessible
[r]References15 refs✓Minimum 10 references required
[w]Words [REQ]2,814✓Minimum 2,000 words for a full research article. Current: 2,814
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18868149
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]43%✗≥80% of references from 2025–2026. Current: 43%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (24 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Fine-Tuned SLMs vs Out-of-the-Box LLMs — Enterprise Cost Reality

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18838660  37stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted33%○≥80% from verified, high-quality sources
[a]DOI7%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed33%○≥80% have metadata indexed
[l]Academic27%○≥80% from journals/conferences/preprints
[f]Free Access33%○≥80% are freely accessible
[r]References15 refs✓Minimum 10 references required
[w]Words [REQ]2,010✓Minimum 2,000 words for a full research article. Current: 2,010
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18838660
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (27 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

The dominant model-selection question in enterprise AI has shifted from "which large language model?" to "should we be using a large language model at all?" This article provides a rigorous economic analysis of fine-tuned small language models (SLMs) versus out-of-the-box large language models (LLMs) for enterprise deployment, drawing on empirical benchmarks from the LoRA Land study, Predibase'...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18838660 37stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted33%○≥80% from verified, high-quality sources
[a]DOI7%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed33%○≥80% have metadata indexed
[l]Academic27%○≥80% from journals/conferences/preprints
[f]Free Access33%○≥80% are freely accessible
[r]References15 refs✓Minimum 10 references required
[w]Words [REQ]2,010✓Minimum 2,000 words for a full research article. Current: 2,010
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18838660
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (27 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

The Small Model Revolution: When 7B Parameters Beat 70B

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18832650  32stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted20%○≥80% from verified, high-quality sources
[a]DOI10%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed20%○≥80% have metadata indexed
[l]Academic10%○≥80% from journals/conferences/preprints
[f]Free Access30%○≥80% are freely accessible
[r]References10 refs✓Minimum 10 references required
[w]Words [REQ]2,208✓Minimum 2,000 words for a full research article. Current: 2,208
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18832650
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]30%✗≥80% of references from 2025–2026. Current: 30%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (19 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

The prevailing assumption in enterprise AI procurement has been that larger models deliver proportionally superior outcomes — that scaling parameters translates linearly into business value. This assumption is wrong, and the evidence in 2026 is now overwhelming. A fine-tuned Phi-3-mini model beat GPT-4o on six of seven financial NLP benchmarks at an inference cost of $0.13 per million tokens ve...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18832650 32stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted20%○≥80% from verified, high-quality sources
[a]DOI10%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed20%○≥80% have metadata indexed
[l]Academic10%○≥80% from journals/conferences/preprints
[f]Free Access30%○≥80% are freely accessible
[r]References10 refs✓Minimum 10 references required
[w]Words [REQ]2,208✓Minimum 2,000 words for a full research article. Current: 2,208
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18832650
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]30%✗≥80% of references from 2025–2026. Current: 30%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams4✓Mermaid architecture/flow diagrams. Current: 4
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (19 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Model Benchmarking for Business — Beyond Academic Metrics

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18827617  15stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted0%○≥80% from verified, high-quality sources
[a]DOI0%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed0%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access0%○≥80% are freely accessible
[r]References12 refs✓Minimum 10 references required
[w]Words [REQ]1,821✗Minimum 2,000 words for a full research article. Current: 1,821
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18827617
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams0○Mermaid architecture/flow diagrams. Current: 0
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (5 × 60%) + Required (2/5 × 30%) + Optional (0/4 × 10%)

Enterprise procurement of large language models (LLMs) continues to rely on academic benchmarks — MMLU, HumanEval, HellaSwag — that were designed for research comparisons rather than business decision-making. This article demonstrates why these metrics systematically mislead enterprise buyers and proposes the Business-Oriented Model Evaluation (BOME) framework, which centres on four operational...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18827617 15stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted0%○≥80% from verified, high-quality sources
[a]DOI0%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed0%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access0%○≥80% are freely accessible
[r]References12 refs✓Minimum 10 references required
[w]Words [REQ]1,821✗Minimum 2,000 words for a full research article. Current: 1,821
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18827617
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]33%✗≥80% of references from 2025–2026. Current: 33%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams0○Mermaid architecture/flow diagrams. Current: 0
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (5 × 60%) + Required (2/5 × 30%) + Optional (0/4 × 10%)
Cost-Effective Ent…Read More
Read more

Autonomous Systems Economics: Replacing Human Labor with Compute

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18822768  22stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted17%○≥80% from verified, high-quality sources
[a]DOI6%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed6%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access11%○≥80% are freely accessible
[r]References18 refs✓Minimum 10 references required
[w]Words [REQ]1,792✗Minimum 2,000 words for a full research article. Current: 1,792
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18822768
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]28%✗≥80% of references from 2025–2026. Current: 28%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams6✓Mermaid architecture/flow diagrams. Current: 6
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (13 × 60%) + Required (2/5 × 30%) + Optional (1/4 × 10%)

The fundamental economic question facing enterprises in 2026 is not whether autonomous systems can replace human labor, but when the compute-labor cost crossover makes replacement economically rational. This article examines the economics of autonomous system deployment across warehouse robotics, transportation, and knowledge work domains. Analysis of real-world implementations reveals that lab...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18822768 22stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted17%○≥80% from verified, high-quality sources
[a]DOI6%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed6%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access11%○≥80% are freely accessible
[r]References18 refs✓Minimum 10 references required
[w]Words [REQ]1,792✗Minimum 2,000 words for a full research article. Current: 1,792
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18822768
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]28%✗≥80% of references from 2025–2026. Current: 28%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams6✓Mermaid architecture/flow diagrams. Current: 6
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (13 × 60%) + Required (2/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Enterprise AI: A Comprehensive Guide to Navigating Complexity and Avoiding the 80% Failure Rate

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18772218  30stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources12%○≥80% from editorially reviewed sources
[t]Trusted20%○≥80% from verified, high-quality sources
[a]DOI2%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed12%○≥80% have metadata indexed
[l]Academic12%○≥80% from journals/conferences/preprints
[f]Free Access4%○≥80% are freely accessible
[r]References51 refs✓Minimum 10 references required
[w]Words [REQ]4,578✓Minimum 2,000 words for a full research article. Current: 4,578
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18772218
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]16%✗≥80% of references from 2025–2026. Current: 16%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams7✓Mermaid architecture/flow diagrams. Current: 7
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (16 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

Executive Summary: Despite unprecedented investment and executive enthusiasm, 80-85% of enterprise AI projects fail to deliver meaningful business value. This comprehensive analysis examines the technical, organizational, and economic factors driving this failure rate, drawing from academic research and industry studies. We present evidence-based frameworks for total cost of ownership (TCO) ana...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18772218 30stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources12%○≥80% from editorially reviewed sources
[t]Trusted20%○≥80% from verified, high-quality sources
[a]DOI2%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed12%○≥80% have metadata indexed
[l]Academic12%○≥80% from journals/conferences/preprints
[f]Free Access4%○≥80% are freely accessible
[r]References51 refs✓Minimum 10 references required
[w]Words [REQ]4,578✓Minimum 2,000 words for a full research article. Current: 4,578
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18772218
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]16%✗≥80% of references from 2025–2026. Current: 16%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams7✓Mermaid architecture/flow diagrams. Current: 7
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (16 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Multi-Provider Strategies: Avoiding Vendor Lock-in While Maximizing Value

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18769559  23stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted0%○≥80% from verified, high-quality sources
[a]DOI0%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed0%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access0%○≥80% are freely accessible
[r]References8 refs○Minimum 10 references required
[w]Words [REQ]2,547✓Minimum 2,000 words for a full research article. Current: 2,547
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18769559
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]38%✗≥80% of references from 2025–2026. Current: 38%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams7✓Mermaid architecture/flow diagrams. Current: 7
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (4 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

Enterprise adoption of large language models (LLMs) has introduced a new dimension of vendor lock-in that differs fundamentally from traditional software dependencies. Unlike switching ERP systems or databases—where migration paths are well-understood—LLM provider transitions involve prompt re-engineering, model behavior differences, and hidden integration costs that can reach six figures even ...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18769559 23stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources0%○≥80% from editorially reviewed sources
[t]Trusted0%○≥80% from verified, high-quality sources
[a]DOI0%○≥80% have a Digital Object Identifier
[b]CrossRef0%○≥80% indexed in CrossRef
[i]Indexed0%○≥80% have metadata indexed
[l]Academic0%○≥80% from journals/conferences/preprints
[f]Free Access0%○≥80% are freely accessible
[r]References8 refs○Minimum 10 references required
[w]Words [REQ]2,547✓Minimum 2,000 words for a full research article. Current: 2,547
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18769559
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]38%✗≥80% of references from 2025–2026. Current: 38%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams7✓Mermaid architecture/flow diagrams. Current: 7
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (4 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more

Specialized vs General Models — When to Use Domain-Specific AI

Posted on by
Applied Research
Applied Research by Oleh Ivchenko  ·  DOI: 10.5281/zenodo.18743173  45stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources8%○≥80% from editorially reviewed sources
[t]Trusted55%○≥80% from verified, high-quality sources
[a]DOI31%○≥80% have a Digital Object Identifier
[b]CrossRef10%○≥80% indexed in CrossRef
[i]Indexed16%○≥80% have metadata indexed
[l]Academic41%○≥80% from journals/conferences/preprints
[f]Free Access47%○≥80% are freely accessible
[r]References51 refs✓Minimum 10 references required
[w]Words [REQ]4,226✓Minimum 2,000 words for a full research article. Current: 4,226
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18743173
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]12%✗≥80% of references from 2025–2026. Current: 12%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams8✓Mermaid architecture/flow diagrams. Current: 8
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (40 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)

Academic Citation: Ivchenko, O. (2026). Specialized vs General Models — When to Use Domain-Specific AI. Cost-Effective Enterprise AI Series. Odessa National Polytechnic University. DOI: 10.5281/zenodo.18746111 Abstract The enterprise AI landscape is undergoing a fundamental shift from general-purpose large language models (LLMs) to domain-specific language models (DSLMs) optimized for particula...

Show moreHide
Applied Research by Oleh Ivchenko DOI: 10.5281/zenodo.18743173 45stabilfr·wdophcgmx
BadgeMetricValueStatusDescription
[s]Reviewed Sources8%○≥80% from editorially reviewed sources
[t]Trusted55%○≥80% from verified, high-quality sources
[a]DOI31%○≥80% have a Digital Object Identifier
[b]CrossRef10%○≥80% indexed in CrossRef
[i]Indexed16%○≥80% have metadata indexed
[l]Academic41%○≥80% from journals/conferences/preprints
[f]Free Access47%○≥80% are freely accessible
[r]References51 refs✓Minimum 10 references required
[w]Words [REQ]4,226✓Minimum 2,000 words for a full research article. Current: 4,226
[d]DOI [REQ]✓✓Zenodo DOI registered for persistent citation. DOI: 10.5281/zenodo.18743173
[o]ORCID [REQ]✓✓Author ORCID verified for academic identity
[p]Peer Reviewed [REQ]—✗Peer reviewed by an assigned reviewer
[h]Freshness [REQ]12%✗≥80% of references from 2025–2026. Current: 12%
[c]Data Charts0○Original data charts from reproducible analysis (min 2). Current: 0
[g]Code—○Source code available on GitHub
[m]Diagrams8✓Mermaid architecture/flow diagrams. Current: 8
[x]Cited by0○Referenced by 0 other hub article(s)
Score = Ref Trust (40 × 60%) + Required (3/5 × 30%) + Optional (1/4 × 10%)
Cost-Effective Ent…Read More
Read more