Email Drafts

Hello RunPod Team,

I'm writing on behalf of the St. Cloud State University Library to inquire about your Academic Research Credits and educational pricing for an upcoming LLM fine-tuning project we are calling "HuskyBot."

Project summary:
  • Institution: St. Cloud State University (SCSU), Minnesota
  • Lead Researcher: Professor Ning Hou
  • Project Coordinator: Gustavo Valdez Paez
  • Goal: Fine-tune an open-weight 7B–8B parameter model (Mistral 7B / Llama 3 8B) on academic corpora (ArXiv, PubMed Central, Semantic Scholar) for a library-facing research assistant
  • Timeline: May 15 – August 15, 2026
  • Project framing: lean academic initiative (under $15K USD total)
  • Estimated GPU need: ~150–250 hours on A100 / H100-class GPUs (LoRA / QLoRA fine-tuning + evaluation)

We would like to apply for your Research Credits program and/or learn about academic discounts and reserved spot pricing that would fit within our budget. Could you share:
  1. Eligibility and application process for the Research Credits program
  2. Best-available academic / reserved pricing for A100 80GB and H100 instances
  3. Whether community-cloud spot instances are recommended for our scale

Happy to provide additional documentation, an institutional letter, or a more detailed compute plan.

Thank you,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Hello Lambda Labs Sales Team,

I'm reaching out on behalf of the St. Cloud State University Library regarding your academic discount program (up to 50%) for our LLM fine-tuning project, "HuskyBot."

Quick overview:
  • Institution: St. Cloud State University, Minnesota
  • Lead Researcher: Professor Ning Hou
  • Project: Fine-tune an open-weight ~7B model on academic data (ArXiv, PubMed Central, Semantic Scholar) and deploy as a library research assistant
  • Compute requirements: ~150–250 GPU-hours on A100 80GB or H100 (LoRA/QLoRA workload)
  • Window: Late May through July 2026
  • Project framing: lean academic initiative (under $15K USD total)

Could you please share:
  1. Application process for the Lambda academic / research program
  2. Best academic rate for on-demand A100 80GB and H100 instances
  3. Whether reserved short-term capacity (2–4 weeks) is available within our budget

We can provide an institutional verification letter on SCSU letterhead and a project summary upon request.

Thank you,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Dear ACM Digital Library Sales Team,

I am writing on behalf of the St. Cloud State University Library to request a tailored pricing proposal for the ACM Digital Library Premium subscription, with potential text-and-data-mining (TDM) use as part of an academic research project, "HuskyBot."

Project context:
  • Institution: St. Cloud State University (SCSU), Minnesota
  • Lead Researcher: Professor Ning Hou
  • Project Coordinator: Gustavo Valdez Paez
  • Use case: Library research assistant + non-commercial fine-tuning of an open-weight LLM on computing literature
  • Project framing: lean academic initiative (under $15K USD); we are evaluating which licensed sources are within reach for Phase 1 versus Phase 2

Could you please share:
  1. The applicable Digital Library Premium tier for SCSU based on our publication output
  2. Multi-year pricing options and any available academic discounts
  3. Whether non-commercial TDM access for LLM fine-tuning falls within standard institutional licensing or requires a separate addendum
  4. Consortia or MnPALS / Minnesota State system pricing options if applicable

We are also happy to be connected with our regional ACM account manager.

Thank you,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Dear IEEE Data Licensing Team,

I am submitting this inquiry on behalf of the St. Cloud State University Library regarding text-and-data-mining (TDM) access to IEEE Xplore content for our non-commercial academic research project, "HuskyBot."

Project summary:
  • Institution: St. Cloud State University, Minnesota
  • Lead Researcher: Professor Ning Hou
  • Use case: Fine-tuning an open-weight LLM (~7B parameters) on academic content, plus a RAG-based research assistant for SCSU students
  • Scope: Non-commercial, internal university use; downloadable model offered to SCSU users
  • Timeline: May – August 2026
  • Project framing: lean academic initiative (under $15K USD)

We would like to understand:
  1. Whether IEEE Xplore Metadata API access is available to SCSU under our current subscription
  2. Cost and process for full-text API access for TDM purposes
  3. Any restrictions specific to LLM fine-tuning on IEEE content
  4. Whether bulk delivery of an IEEE content subset is available within our budget

I will also submit the IEEE Xplore Data Licensing Options form for tracking. Please feel free to redirect this to the appropriate licensing specialist.

Thank you,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Dear Elsevier Team,

I am writing on behalf of the St. Cloud State University Library regarding non-commercial Text-and-Data-Mining (TDM) access to ScienceDirect via the Elsevier Developer Portal for our academic research project, "HuskyBot."

Project summary:
  • Institution: St. Cloud State University (SCSU), Minnesota
  • Lead Researcher: Professor Ning Hou
  • Project Coordinator: Gustavo Valdez Paez
  • Use case: Non-commercial fine-tuning of an open-weight LLM on academic literature; output used in a library-facing research assistant for SCSU students
  • Timeline: May – August 2026
  • Project framing: lean academic initiative (under $15K USD)

We would like clarification on:
  1. Whether SCSU's existing institutional subscription qualifies for the standard non-commercial TDM API key on the Elsevier Developer Portal
  2. Whether LLM fine-tuning on the ScienceDirect dataset is permitted under the standard TDM agreement, or requires a special addendum
  3. Documentation requirements for compliance (deletion of corpus on completion, retention rules for model outputs, snippet redistribution limits)
  4. Recommended point of contact / our SCSU Elsevier account manager

We can provide additional project documentation upon request and are committed to full compliance with Elsevier's TDM policies.

Thank you,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Hello Semantic Scholar Team,

I'd like to request an elevated API key and access to the S2ORC full-text dataset for an academic, non-commercial LLM fine-tuning project at St. Cloud State University, called "HuskyBot."

Quick details:
  • Institution: St. Cloud State University, Minnesota
  • Lead Researcher: Professor Ning Hou
  • Project: Fine-tune an open-weight ~7B model on high-quality academic corpora and deploy as a library research assistant
  • Use case: Non-commercial, university research; downloadable model offered to SCSU users
  • Timeline: May – August 2026

Could you please advise on:
  1. The application process for S2ORC bulk dataset access
  2. Rate-limit upgrades for the Semantic Scholar Academic Graph API for our use case
  3. Any attribution / licensing requirements we should be aware of for derivative LLMs

Thank you for the incredible resource you provide to the academic community.

Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library
[[email protected]]

Hello [Name],

Professor Ning Hou and I are kicking off the "HuskyBot" initiative — a custom AI research assistant and writing helper for SCSU Library users — and I'd like to coordinate API access to a few of our existing systems.

What I'd like to discuss (Phase 2 — once Phase 1 is functional):
  1. Library catalog / discovery layer API access (read-only)
  2. Authenticated proxy / API access to subscribed databases (EBSCOhost, ProQuest, JSTOR, ScienceDirect, etc.) where licensing permits
  3. SSO / Shibboleth integration for student-facing access
  4. Hosting options on SCSU infrastructure for the public web interface

The project is summer-semester scoped (May 15 – August 15, 2026) with a modest budget covering compute and tooling. Phase 1 uses only open / public sources (ArXiv, PubMed Central, Semantic Scholar) and does not require any internal access — but I want to start the conversation now so Phase 2 integration is ready when the model is.

Would you have 30 minutes for a kickoff call this week or next?

Best,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library

Dear [Dean / Director Name],

Professor Ning Hou (Lead Researcher) and I are launching "HuskyBot," a summer-semester initiative to build a custom AI research assistant for the SCSU Library. I'd like to share a brief project overview and request your support.

In one paragraph: HuskyBot is a hybrid library chatbot and academic writing assistant, powered by an open-weight LLM that we will fine-tune on high-quality academic corpora (ArXiv, PubMed Central, Semantic Scholar). It will be available both as a web app on the library site and as a downloadable model students can run locally. The full project runs May 15 – August 15, 2026, on a modest budget that must be committed before June 1.

What we need from your office:
  • Endorsement of the initiative as an officially-recognized library project
  • Introductions to (a) library systems / IT, and (b) anyone who manages our database vendor relationships (Elsevier, IEEE, ACM, EBSCO, etc.)
  • Guidance on procurement so allocated funds can be committed against GPU compute and licensed dataset access before the June fiscal deadline

I've put together a public project site that walks through goals, timeline, budget, and technical approach — happy to share the link and meet for 30 minutes at your convenience.

Thank you for your support,
Gustavo Valdez Paez
Lead Developer & Project Coordinator — HuskyBot
St. Cloud State University Library