Evan Frisch

Rather than treating the student information system as the whole solution, the posts show how a small tool can connect it to other data or fill a missing feature.

Boundary Finder uses a district boundary file and map data to identify assigned schools for a new address that has not yet reached the SIS. They also describe a hall pass app built after a district could not customize the feature in its SIS.

Read original · Original 2

That lets a district work with outside organizations without giving them student rosters.

In the volunteer-hours prototype, a bridge table links district student records to separate volunteer records.

• Community coordinators see a chosen display name and internal volunteer ID, while the district keeps the link and audits totals.
• The post says the student and district-administrator views were described but not built in the demo.

Read original

Automation should keep information current, not erase the working record or decide that every mismatch is an error.

Grad Tracker refreshes graduation-requirement columns in a Google Sheet while preserving staff notes about open items. Boundary Finder flags enrollments that differ from geographic boundaries for staff review; they note that an out-of-boundary placement may be legitimate.

Read original · Original 2

The goal is to produce a report and reuse a method for connecting to the next source.

District Insight draws on two Census Bureau APIs to compare a district with peers and the state, then creates a presentation for staff to review and customize. They describe a repeatable process: find endpoints, verify variable names, add retries and caching, and translate field names into readable ones.

Read original

Small, maintainable implementations matter because district technical staff should be able to host and change them as needs shift.

They describe prototypes built with compact stacks, including a Python backend and SQLite, and argue that the right training helps district data managers, SIS administrators, and IT staff turn their knowledge into working tools.

Read original · Original 2

A district asked for a change to the hall pass feature in its student information system, the SIS could not support it, and Frisch built a working alternative in about six hours. For newcomers wondering whether the posts describe real software or general commentary, this is the place to start.

In the demo, a teacher issues a digital hall pass, choosing a destination and duration.

• The pass appears on a hall monitor's dashboard within five seconds, with no manual refresh.
• Staff on hall duty can look up any student to see whether a pass is valid, overdue, or nonexistent, and check students back in when they return.
• The dashboard lists every student with an outstanding pass, flagging overdue ones and showing departure times, destinations, and who issued each pass.
• Avatars in the demo are placeholders for student photos a real deployment would pull from the district's own storage.
• A real version would sync rosters from the SIS, use Google or Microsoft single sign-on, and set permissions by role: teachers issue passes to their students and can verify any student's pass; hall duty staff verify but cannot issue; administrators do both and configure the system.
• The stack is deliberately small — FastAPI, HTMX, SQLite, one Python process serving HTML, no JavaScript framework or build step — so district technical staff could host and maintain it themselves.

For brands planning a campaign

For a K-12 edtech vendor, this is a concrete picture of the integration friction districts face when a bundled SIS feature cannot match local policy. It supports collaboration or content angles around tools that extend and complement the SIS rather than replace it, aimed at the moment a district is told, "the system doesn't do that."

It is the set's most transferable design idea.

Many high schools track student volunteer hours at community organizations, sometimes as a graduation requirement, sometimes for recognition at graduation.

• Students earn credit for hours logged with outside organizations, but student records are protected and community partners should not hold student rosters.
• In the demo, a community organization coordinator signs in, generates a one-time enrollment code for a new student volunteer, logs hours against a linked volunteer record, and confirms hours volunteers logged themselves.
• A dashboard shows the confirmed total rising once both sides confirm.
• The design uses a bridge table to connect students on the district side, with real identifiers, to volunteer records on the organization side, which carry none.
• Coordinators see only a chosen display name and an internal volunteer ID.
• Hours count only after the organization and student confirm them, and the district owns the bridge and audits the totals.
• In a real deployment, students would sign in with school single sign-on and claim codes from approved organizations, coordinators would use their own credentials, and district administrators would approve organizations, audit hours, and export verified totals for graduation clearance.
• Frisch built it in about seven hours with Claude Code.
• The demo details only the coordinator's view; the student and district administrator sides are described in narration but not built.
• The stack is server-rendered HTML, one Python backend, SQLite, and light in-page interactivity, with no frontend framework.

For brands planning a campaign

A vendor whose products handle student data shared with outside partners can use this as a reference pattern for privacy-preserving exchange: identity stays on one side, work is recorded on the other, and nothing counts without confirmation from both.

• It also offers a content angle: explain FERPA-aware design through a working loop.
• Ask which narrated features are built; the demo does not include them all.

Newcomers can see the pattern in one readable example.

A family moves into a new house on a new street and registers at the district office.

• The address is not in the student information system, the street is too new for the SIS address table, and the assigned elementary, middle, and high schools are undefined.
• The district already maintains boundary lines as a file, and the new road already appears in map data, so the information is available, just outside the system in use.
• The tool geocodes any typed address, runs a point-in-polygon test against the district boundary file, and returns the assigned schools.
• If the address is already in the SIS, it confirms what the SIS says; if not, it uses current map data and the boundary file.
• A batch view shows every pending registration with status badges, letting staff triage a morning's stack of new families on one screen.
• An enrollment overlay plots existing students on a map and flags cases where the school recorded in the SIS does not match the geographic boundary.
• Frisch notes that out-of-boundary enrollments can be legitimate and documented, so the tool surfaces cases for staff judgment rather than correcting them automatically.
• It is a small Flask app reading a GeoJSON boundary file and CSV student extracts, demonstrated with a fictional central Indiana district and fully synthetic student data.
• It was built in about four hours with Claude Code.

For brands planning a campaign

Registration and enrollment are concrete buyer problems. This prototype shows how to layer geographic data over records a district already holds, making it a fit for content that demonstrates district data integration.

The constraint is the interesting part.

Every spring, high schools check whether seniors have met every graduation requirement — transcript credits, outstanding fees, cap and gown, library books returned, district device returned.

• The information lives across different systems or screens in the SIS, and the deadline does not move.
• Staff commonly build a Google Sheet with one row per senior, one column per requirement, a status column reading Eligible or Open Items, and notes such as "Returned book to library on 5/20, awaiting confirmation" or "Parent contacted; check arriving by mail."
• Those notes become the record of work in progress.
• Frisch built Grad Tracker so a script reads the senior class, computes status against each requirement, and writes results to a Google Sheet, merging on student ID: requirement columns refresh every run while the notes column is preserved from the existing sheet.
• The demo reads from a single CSV, but the same approach could pull transcript and fee data from the SIS, book status from the library system, and device assignments from inventory, then sync everything to one sheet.
• Frisch points to the same pattern for new student enrollment, device assignments, repair tracking, athletic eligibility, and food service balances.
• The video shows the script run manually; a district would schedule it hourly, nightly, or on whatever cadence it needs.
• It is a Python script writing to Google Sheets through the Sheets API, built in about four hours with Claude Code using fully synthetic data.

For brands planning a campaign

This names an integration requirement vendors often miss: write-back that preserves human annotations instead of overwriting them. It supports content about working with spreadsheets staff already trust rather than demanding they abandon them.

A superintendent wants a presentation comparing the district to similar districts on poverty rates, household income, language demographics, and educational attainment.

• The Census Bureau publishes that data for every district in the country, free and regularly updated.
• But data outside a district's own reports and queries often arrives only through APIs — federal datasets, assessment platforms publishing scores through a web service, state accountability portals, an LMS exposing assignment-level detail.
• That means API keys, authentication, endpoints, JSON response formats, pagination, and rate limits.
• For most district technical staff, that has meant handing the job to a vendor or downloading CSVs and charting them by hand.
• District Insight, built as a proof of concept, pulls from two Census Bureau APIs, identifies peer districts by population and poverty rate, compares the district to peers and the state average across five measures, generates a written narrative with an AI model, and produces a downloadable nine-slide presentation with charts, a comparison table, and key takeaways.
• Frisch calls the output a starting point for staff to review and customize.
• A district building this for itself would scope it more narrowly to its own district and a handful of known peers, making it simpler and faster.
• The more important outcome was capturing the pattern: discover the endpoints, verify the actual variable names, add retry logic, cache responses, and map cryptic field names to readable ones.
• That pattern was saved as a reusable Claude Code skill, so later integrations start with the approach in place.
• The API client itself took about three minutes to generate; the full project took under eight hours across several sessions, most of it spent on presentation design and refinements.

For brands planning a campaign

Relevant to any vendor whose value depends on districts pulling in outside data, and a natural co-marketing angle with a district-facing customer around practical API integration. The narrative is model-generated and presented as a draft for staff review, not an authoritative analysis.

The five prototypes are Boundary Finder (address-to-school assignment using a boundary file), a digital hall pass app, a volunteer-hour tracker for community-service credit, Grad Tracker (senior graduation requirements), and District Insight (peer-district comparisons).

• Builds take roughly four to eight hours each.
• Frisch makes the case by showing the tool running, not describing what a vendor could build.

Build a working proof of concept, record the demo, and state the build time and scope shown.

The demos have clear limits: the volunteer tracker shows only the coordinator's view; the student and district-admin sides are described in narration, not built. The hall pass demo uses illustrated avatars in place of real student photos, and Grad Tracker and Boundary Finder run on synthetic data. Ask to see the student and district-admin views before assessing the volunteer tracker.

For brands planning a campaign

A K-12 edtech company can evaluate claims by inspecting working software, not a written pitch. The natural collaboration is a sponsored build or build-along around a named district workflow, using the same video format—not a testimonial or general-topic talk.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5

Community coordinators see only a chosen display name and an internal volunteer ID, never a student roster.

• Hours count only after both the organization and the student confirm them; the district owns the bridge and audits totals.
• The post argues that district staff already understand what counts as student PII and what FERPA requires, but their tools make it difficult to put those principles into practice.
• Other demos use synthetic data for the same reason.

Separate identifiers by design: one side holds PII, the other holds none, and the district controls the link and the audit.

For brands planning a campaign

For content aimed at district buyers, the test is whether it shows the mechanism—who holds identifiers and what an outside partner can see—not just compliance language. A vendor whose product touches student data can sponsor a build demonstrating its integration under FERPA-shaped constraints, then be evaluated on that demonstration.

Examples: Source 1 · Source 2 · Source 3

One district asked about customizing the hall pass feature in its student information system, but the SIS could not support the changes.

• A new street was not yet in the SIS address table, leaving the assigned schools undefined for a family registering that day.
• Graduation-requirement checks are assembled by hand across transcript credits, fees, library books, and device returns.
• Volunteer hours are earned at outside organizations.
• In each case, the information exists but sits outside the system the staff member is using.
• The posts consistently identify the bottleneck not as staff understanding of the work, but as the gap between knowing what should exist and being able to build it.

For brands planning a campaign

The strongest fit is a complementary role alongside the SIS, not a replacement pitch.

• Vendors that expose data, document their schema, or supply an API can put their product inside a working demo of a district workflow.
• The seam between systems is where this audience already looks for value.

Examples: Source 1 · Source 2 · Source 3 · Source 4

Boundary Finder is a small Flask app that reads a GeoJSON boundary file and CSV extracts. The stated reason is ownership: a district's technical staff could host, modify, and maintain the result without outside help each time a policy changes, in days instead of months.

For brands planning a campaign

Platform, hosting, and identity vendors can be assessed as components within these builds.

• Google or Microsoft SSO for staff authentication is named as the path to real deployment, and permissions are described by staff role.
• Because the audience includes the staff member who will own and modify the tool after launch, maintenance burden is part of the buying decision.

Examples: Source 1 · Source 2 · Source 3

District Insight pulls from two Census APIs, identifies peer districts by population and poverty rate, compares five measures against peers and the state average, generates a written narrative, and produces a nine-slide presentation. The post identifies the mechanics—API keys, authentication, endpoints, JSON formats, pagination, and rate limits—as the deterrent, leading districts to hand the work to a vendor or build charts by hand.

The repeatable integration pattern is to discover endpoints, verify actual variable names, add retry logic, cache responses, and map cryptic field names to readable ones.

• Frisch saved the pattern as a reusable skill so the next integration starts from it.
• The API client took about three minutes to generate; most of the eight hours went to presentation design.

For brands planning a campaign

API-first edtech and data providers have the clearest angle: a documented API paired with an integration build is directly on-topic.

• A sponsored session connecting a vendor's API to a real district workflow fits both the subject matter and video format.
• The reusable-skill framing lets the vendor appear as one integration among several.

Examples: Source 1 · Source 2

It rewrites requirement columns while preserving the notes column, where staff record in-progress work such as a returned book awaiting confirmation.

• The stated opportunity is to keep the sheet current on a schedule without overwriting those notes.
• The same pattern extends to enrollment, device assignments, repair tracking, athletic eligibility, and food service balances.
• Boundary Finder adds a batch view for triaging a morning's pending registrations on one screen, plus an enrollment overlay that flags cases where the school in the SIS does not match the geographic boundary—so staff can judge whether an out-of-boundary enrollment is warranted.

Merge on a stable key, update machine-owned fields, leave human-written notes untouched, and flag mismatches for a person to decide rather than resolving them automatically.

For brands planning a campaign

Workflow and data-sync vendors should show that human notes and judgment survive automation. A campaign asking district staff to abandon their spreadsheets and habits wholesale misses the point; one that carries their work forward fits it.

Examples: Source 1 · Source 2

Two posts end with the line that the right training is what makes the difference. The prototype series demonstrates what that training unlocks rather than announcing a product.

For brands planning a campaign

The offer is enablement, not tool procurement.

• A vendor's credibility depends on whether its training or developer experience leaves district staff able to build and maintain their own tools.
• Programs focused on capacity building—workshops, build challenges, documented APIs with tutorials—fit the demonstrated priorities better than license-led messaging.

Examples: Source 1 · Source 2 · Source 3

The creator reports build times of roughly three to eight hours per project and credits the tool with closing the gap between district staff knowing what should exist and being able to build it. The write-ups do not state whether there is a sponsorship, partner status, or commercial agreement with Anthropic; ask for disclosure if that matters to a campaign.

View example

The prototypes use Google as an available district platform.

View example

View example

Chernikoff's posts explore why districts keep or drop a tool and what founders get wrong, including a talk with a district superintendent about moving on from a solution that is not creating value, and a startup challenge where Chernikoff coaches education founders. Frisch approaches the same market conversation from the technical end: what a district can build and where existing systems fall short.

Chernikoff advises and coaches edtech founders on sales, procurement, and retention, drawing on other leaders' views and their own programs. Frisch shows running software through build times and demo walkthroughs, and argues for district staff capacity rather than founder strategy.

Read original · Original 2

Terry writes about teacher evaluation ratings that vary by campus and appraiser, the evidence behind scores in high-stakes Texas incentive models, and vendors needing to work within district leaders' constraints. Frisch's posts show how software can fit district policy and data realities.

Terry focuses on teacher evaluation, rating consistency, and evidence quality, promoting a product and advisory role in that space. Frisch focuses on district data workflows and self-built tools, demonstrating prototypes rather than advocating for a commercial platform.

Read original · Original 2

Makani tells the story of a nonverbal student using a speech board in church, runs live build sessions with teachers and speech-language pathologists, and answers the special education director's question about justifying another tool by counting specialist hours. Frisch also argues from a specific district problem and a demonstrated tool rather than product claims.

Makani's posts make adoption and return-on-investment arguments for a commercial assistive-technology product aimed at special education teams, using customer moments and staffing math. Frisch builds proof-of-concept tools for general district data workflows, reports build times and technical design, and makes no product offer.

Read original · Original 2

Chernikoff's posts take the buying cycle itself as their subject.

• One argues that winning a contract is not enough: a district should be able to move on from a solution that stops delivering value.
• It frames this as a lesson for education founders.
• The other shows sponsorship in practice: lead sponsor of the Dragon's Den EdTech Startup Challenge at the Canadian EdTech Leadership Summit, with coaching for applicants before the event, private preparation with six finalists, and a prize of $40,000+ in coaching, tools, visibility, and a full year inside the EdSales Revenue Machine.
• A vendor seeking founder-to-founder or education-sales credibility, or a packaged sponsorship with coaching attached, gets a closer match here than with Frisch.

The tradeoff is audience and subject.

• Chernikoff writes to sellers about the procurement cycle; Frisch writes to data managers, SIS administrators and IT staff, showing software running against district files.
• A campaign that needs district technical staff to recognise an integration in their own data processes will not get that from a sales-motion audience, and Chernikoff's posts show no product or integration.
• Treat the challenge sponsorship as evidence that the format exists, not as a template for API-led content.
• The two posts available are dated 20 June 2026 and 8 October 2026.

Examples: Source 1 · Source 2

Terry approaches the problem from two directions.

• One post points to the gap between what evaluation systems record and what drives teacher growth.
• It argues that buying a product does not guarantee transformation unless the vendor helps the district through implementation, with the margin for error in TIA-aligned models described as narrow as 0.2%.
• The other names the variability directly: two principals, the same lesson, different ratings, because the difference lies in interpretation.
• Compensation decisions then shift, TIA designations change and trust erodes.
• The ask is confidence that every rating rests on aligned, observable evidence before decisions are finalised.
• Terry also discloses a position inside the content as President of the RefynED Advisory Board, making these posts a working example of a disclosed advisory relationship.
• A vendor whose sale depends on evaluation evidence holding up under challenge, or on teacher retention, is better served here.

The tradeoff is scope.

• This is a narrow, high-stakes process that does not touch the integration, API or data-plumbing problems Frisch makes the subject of their posts.
• Because the advisory relationship sits inside the argument, the content reads as advocacy for a specific tool rather than a neutral comparison of options.
• It is useful as a disclosed model, but not as an independent evaluation.
• A campaign that needs district staff to see a product working inside a spreadsheet or beside the SIS fits Frisch's build-and-walkthrough format better.
• The two posts available are dated 6 February 2026 and 22 April 2026.

Examples: Source 1 · Source 2

The two posts do different jobs.

• The first tells a story with an outcome: a boy at a church in Alabama who does not speak, a speech therapist building him a five-verse board in Glint, and the congregation hearing him answer for the first time.
• It follows with live board-building with teachers and SLPs at AETC in Huntsville, where boards were made on the spot and sent to phones by QR code before people left their seats.
• The second offers a discovery method.
• Instead of pitching, Makani asks a director how many staff they have, how many students per specialist, how many hours of planning time per week, and how many hours are worked off the clock.
• Then Makani lets the arithmetic carry the case: five unpaid hours a week cut to thirty minutes, described as buying back mental health, retention and recruiting in a profession with a 30% shortage.
• For a vendor selling into special education, this is a complete demand-generation argument.

The tradeoff is audience width and tone.

• The readership is special education and assistive technology, so a district data-process product would reach one part of the district rather than the data managers and IT staff Frisch addresses.
• The second post is closer to a live sales conversation than a demonstration: it describes the opening minutes of a call and the pitch that follows.
• A vendor wanting to teach a technical audience through inspectable screens would be borrowing a persuasion format rather than a build format.
• Frisch shows code running and states what they stubbed; Makani shows a use case and a way of framing value.
• The two posts available are dated 25 June 2026 and 2 June 2026.

Examples: Source 1 · Source 2