AI for Genealogy: Civil War Widow Denied - AI Uncovers Heartbreaking Pension Files

She had six children, a 33-year marriage, and witnesses who knew her as Mrs. Hoffman. So why couldn't this Civil War widow prove she was married?

In this heartbreaking case study, we dive into a 173-page pension file that reveals how one missing piece of paper destroyed a family—and how FREE AI tools helped me uncover the devastating truth behind rejected widow pension applications.

What You'll Discover in This Episode:

✅ The Shocking Reality of Civil War Pensions - Why thousands of widows were denied benefits despite being legally entitled (this will make you angry)

✅ AI Handwriting Transcription That Actually Works - I'll show you the exact FREE tool that transcribed desperate letters written in shaky pencil on cheap paper—Catherine's own words: "My children are hungry"

✅ Free Legal Research at Your Fingertips - Learn how Perplexity helped me understand complex 1890s pension laws in 15 seconds (no law degree required)

✅ Multi-Document Analysis Like a Pro - Watch Claude compare witness statements, legal documents, and official records to spot exactly what the pension examiner would have seen

✅ The Church Fire That Changed Everything - How destroyed records created an impossible bureaucratic nightmare (and why this happened to thousands of families)

Copy-Paste Ready Prompts Included: Every AI technique demonstrated with exact prompts you can use today on your own genealogy research—all with 100% FREE tools (no subscriptions needed)

The Unresolved Mystery: Catherine Hoffman never got her pension. After her husband Samuel survived three years of Civil War combat, she spent 24 years dependent on her children. The farm was sold for unpaid taxes. But there's a fourth application mentioned in the file that I CAN'T FIND...and I need your help.

Perfect For:

• Family historians researching Civil War ancestors
• Anyone stuck with rejected applications (pension, land, citizenship)
• Genealogists who want to use AI ethically and effectively
• Researchers dealing with destroyed church records or missing documents
• Anyone who's ever hit a brick wall with bureaucratic requirements

AI Tools Featured (All FREE):

• Perplexity (5 searches/day free tier)
• Gemini via Google AI Studio (completely free)
• Claude (generous free tier)

What Makes This Episode Different: This isn't just another genealogy tutorial. It's a real mystery with emotional stakes, showing you techniques that work on ANY rejected application—pension, land claim, citizenship, you name it. Plus, you'll learn why verification matters just as much as discovery (Golden Rule: AI is your research assistant, not your researcher).

Keywords: Civil War genealogy, widow pension research, AI genealogy tools, handwriting transcription free, genealogy brick wall, Pennsylvania German ancestors, destroyed church records, rejected pension application, genealogy case study, family history research, AI for genealogy, Civil War pension files, genealogy podcast 2025

Connect with Ancestors and Algorithms:

📧 Email: ancestorsandai@gmail.com
🌐 Website: https://ancestorsandai.com/
📘 Facebook Group: Ancestors and Algorithms: AI for Genealogy - www.facebook.com/groups/ancestorsandalgorithms/

Golden Rule Reminder: AI is your research assistant, not your researcher.

Join our Facebook group to share your AI genealogy breakthroughs, ask questions, and connect with fellow family historians who are embracing the future of genealogy research!

New episodes every Tuesday. Subscribe so you never miss the latest AI tools and techniques for family history research.

Transcript

So, there was a moment when I was reading through a pension file, one of those massive 200-page Civil War widow's pension applications, and I found a letter that made me stop cold. It was written in pencil, on cheap paper, by a woman named Catherine Hoffman. The handwriting was shaky, like she'd written it late at night by candlelight. And the first line said, I am writing to you because I have no other choice. My children are hungry. That was in 1892, 27 years after the war ended. And this widow of a Union soldier, a man who'd served honorably, who survived three years of combat, couldn't get her pension. Not because he hadn't served, not because she wasn't his widow. But because she couldn't prove they were married. The pension office had rejected her application twice. Insufficient proof of marriage, they said. And here's the thing that hooked me. 

Catherine knew she had witnesses. She had children. She'd been Mrs. Catherine Hoffman for over 30 years. So, why couldn't she prove it? And that question sent me down a rabbit hole that would show me a rabbit hole that would show me exactly how AI can help us not just find documents, but understand the devastating bureaucratic nightmares our ancestors faced. Because this isn't just a genealogy mystery, it's a window into how one missing piece of paper could destroy a family. Welcome to Ancestors and Algorithms, where family history meets artificial intelligence. I'm your host, Brian, and today we're diving into one of the most heartbreaking types of genealogy research. Civil War Widow Pension Applications. In previous episodes, we explored how AI can help with property I'm a wizarding mystery mysteries. Today, a widow fighting for survival benefits while navigating a legal system that seemed designed to deny her. This is a case study episode, which means we're going to walk through this entire research challenge together, step by step by step. You'll see the exact AI prompts I used, the techniques that worked, the dead ends I hit, and how multiple AI tools, all free ones by the way, helped me piece together Catherine's story. And fair warning, this one doesn't have a neat tidy ending, but sometimes the most valuable genealogy lessons come from the unresolved mysteries. Let me back up and tell you how I found Catherine. I was researching Pennsylvania Civil War soldiers from Berks County. This was about six months ago, and I stumbled across a pension file index card for a soldier named Samuel Hoffman. He had served in Company K, 93rd Pennsylvania Infantry, enlisted in 1861, mustered out in 1864, honorably discharged, all pretty straightforward. But what caught my eye was the notation on the index card. Widow's application, rejected, 1892. See file for details. Now, rejected pension applications are actually gold mines for genealogists. Because when an application gets denied, the widow usually fights back. She gathers more evidence. She finds more witnesses. She writes letters. All of which means there's often more documentation in a rejected application than in an approved one. So, I ordered the pension file from the National Archives. When it arrived, all 173 pages of it, I realized this wasn't just a rejected application. It was a tragedy. Here's what I pieced together from the initial documents. Samuel Hoffman married Catherine Schneider in 1858 in a small rural community outside of Redding, Pennsylvania. He was a farmer. She was the daughter of German immigrants. They had their first child in 1859, a son they named Jacob. When the war came, Samuel enlisted. He fought at Fredericksburg, Chancellorsville, and Gettysburg. He came home in 1864, one of the lucky ones who survived, and went back to farming. But the war had broken something in Samuel. The pension file includes a statement from a neighbor saying he was never the same. He had trouble sleeping. He had trouble sleeping. He startled at loud noises. What we'd call PTSD today, they called soldier's heart, or just said he was melancholic. Samuel died in 1891 at age 54. Heart failure, the death certificate said. He'd never applied for his own veteran's pension. Catherine was 49 years old with six children, the youngest just 12. The farm was mortgaged. She had no income. So, in January 1892, she applied for her widow's pension. And that's when everything fell apart. The pension office sent back her application with a one-line rejection. Insufficient proof of legal marriage to the soldier. Now, you might be thinking, wait, they had six children together. They were married for over 30 years. How hard could it be to prove they were married? Here's the problem. Catherine and Samuel were married in a small German Reformed church in 1858. The ceremony was conducted in German. The pastor recorded it in the church register. But in 1889, just two years before Samuel died, that church burned down. The wooden structure was old. A fire started in the stove one winter night. The whole building was gone by morning, and with it, the church records. Bath records, birth records, marriage records, death records, all of it, gone. Catherine had no marriage certificate. There was no county record. Pennsylvania didn't require civil marriage registration until 1885. And even then, it wasn't consistently enforced in rural areas. She had her memory. She had her children. She had neighbors who'd known them as husband and wife for decades. But she didn't have the one thing the pension office demanded, documentary proof. And this is where I realized that understanding Catherine's situation required understanding not just what happened, but why the pension system worked the way it did. Because there's a real person. Because there's a whole legal and historical context here that changed how I read every document in this file. Remember our golden rule. AI is your research assistant, not your researcher. And in this case, AI was about to help me understand a complex historical situation that I couldn't have wrapped my head around on my own. So, here's what I did. And I want you to follow along, because the techniques I'm about to show you work for any genealogy situation, where you're trying to understand historical legal requirements, bureaucratic processes, or why your ancestors faced the obstacles they did. The first thing I needed to figure out was, what exactly did the pension office require for proof of marriage in 1892? And why were they so strict about it? Now, I could have spent hours, maybe days, going through books about civil war pensions, trying to understand the legal requirements. And honestly, that's what I would have done before AI. But I knew perplexity excels at exactly this kind of historical research question. It searches the internet, pulls from authoritative sources, and cites everything. So, I can see exactly where the information is coming from. I opened Perplexity, the free version, which gives you five searches a day, more than enough for most genealogy research. I asked it this, quote, what were the legal requirements for proof of marriage and civil war widow pension applications between 1862 and 1900? What happened when church records were destroyed, end quote? Now, notice how I structured that prompt. I gave it the specific topic, proof of marriage requirements, the specific context, civil war widow pensions, the timeframe, 1862 to 1900, the complication, destroyed church records. Within about 15 seconds, perplexity came back with a detailed answer citing multiple sources. And here's what I learned. The pension office did recognize common law marriages. Congress had actually passed a law in 1882 that said marriages should be proven according to the law of the place where the parties resided, which meant Pennsylvania state law applied. And Pennsylvania did recognize common law marriage. If a couple cohabitated and held themselves out as husband and wife, that was legally valid. So, why was Catherine being rejected? Here's where it gets interesting. Perplexity sources showed me that while common law marriage was legally valid, the pension office had a long history, going back to the Revolutionary War, of requiring more proof than the courts would require. The pension examiners wanted either a marriage certificate or multiple witness affidavits from people who could testify about the marriage. And not just any witnesses. They wanted people who were present at the wedding, or who could provide specific details about the relationship over time. This completely changed how I understood Catherine's situation. She wasn't being rejected because she wasn't legally married. She was being rejected because she couldn't provide the type of proof the pension office demanded. And here's what made me realize AI could help even more. In the pension file, there were handwritten letters from Catherine trying to explain her situation. But the handwriting was... well, let me just say that 19th century cursive written by someone with limited education using a dull pencil on cheap paper that's been sitting in a file for 130 years? It's not easy to read. I've looked at these letters before and given up trying to transcribe them. Too hard. Too time-consuming. But now I knew that understanding Catherine's own words might be the key to understanding what happened. This is where Gemini comes in, and specifically Google AI Studio, which is the free interface that gives you access to Gemini 3. Now, here's something important I've learned. If you go to the regular Gemini chatbot and try to transcribe a handwritten document, you'll get decent results. But, if you use Google AI Studio, which is completely free, you just need a Google account, the accuracy is dramatically better. I pulled up Google AI Studio, uploaded a photo I'd taken of Catherine's first letter, the one that started with "my children are hungry" and I gave Gemini this very simple prompt: "Transcribe this handwritten letter and keep line breaks" That's it. I didn't need to explain that it was from 1892 or that the handwriting was challenging or anything else. "Just...transcribe and keep line breaks" And within maybe 30 seconds, Gemini gave me this: "Berks County, Pennsylvania, March 15th, 1892" "To the pension office: "I am writing to you because I have no other choice" "My children are hungry, and I cannot pay the mortgage on our farm" "My husband, Samuel Hoffman, served faithfully in the war, and I am his widow, and I am asking for what is owed to me" "You say I must prove we were married, but our church records burned" I can bring you neighbors who knew us as man and wife for 33 years" "I can show you our children: all six of them" "And Samuel's name is written on every one of their birth records as father" do not understand what more proof you need" "Everyone in our community knows we were married" "Pastor Mueller married us in 1858, and he is dead now, but his son remembers it" "Please, I am begging you to help me" "Mrs. Kathryn Hoffman" "I'm not going to lie, reading that transcription hit me hard" "This wasn't just a genealogical puzzle" "This was a real woman, desperate" "Watching her family fall apart because of bureaucratic requirements she couldn't meet" "But here's what the A. I. transcription gave me" "Specific details I could verify" "She mentioned Pastor Mueller" "She mentioned 33 years of marriage, which would date the marriage to 1859" "No, she said they were married for 33 years by 1892, which would be 1858 or early 1859" "She mentioned six children" "She mentioned the farm mortgage" "Every one of those details was something I could potentially verify with other documents" "Now, I had a clear transcription of Kathryn's letter, but I needed to understand what she was up against legally" "What specific evidence, could she provide that the pension office would accept" "I took the perplexity research I'd done and Kathryn's letter and I fed both into Claude" "Claude is excellent at analyzing complex documents and understanding legal requirements" "The free version of Claude gives you plenty of daily messages to work with for genealogy research" "Here's the exact prompt I used" "I'm researching a Civil War widow's pension application from 1892" "The widow's marriage to the soldier is unquestioned" "They were married for 33 years and had six children, but the church records burned" "Based on the historical legal requirements for a widow's pensions, what types of alternative evidence could she provide that the pension office would accept" "Be specific about what witness affidavits needed to include" "I also uploaded Kathryn's letter and the pension office rejection letter" "Claude came back with something really valuable" "It analyzed both the historical legal requirements and Kathryn's specific situation" "It gave me a breakdown of what she needed" "Number one: witness affidavits from people who knew them as husband and wife over time" "Not just people who remembered the wedding, but people who could testify to their cohabitation and public presentation as married" "Number two: evidence of Samuel claiming Kathryn as his wife on official documents" "Things like land deeds, military records, tax records, where he listed her as his spouse" "Number three: children's birth records" "But here's the key" "The records needed to show Samuel acknowledging paternity" "Not just his name listed as Father" "Number four: evidence from community members" "Statements from church members, neighbors, anyone who could testify they were known as Mr. and Mrs. Hoffman" "Claude also pointed out something I'd missed" "The pension file rejection letter mentioned that Kathryn's first submission only included two witness statements and both were from her adult him. That was the problem. The pension office didn't accept testimony from the children because they had a vested interest in the pension being approved. She needed witnesses with no financial stake in the outcome. Armed with this understanding, I went back through the 173-page pension file looking for the supporting documents Catherine submitted in her second application attempt. And this is where having AI as a research assistant becomes invaluable. Because I was now looking for specific things, affidavits from non-family members, official records showing Samuel's acknowledgement of Catherine as his wife, anything that fit the requirements Claude had identified. The file was organized chronologically, but not in a way that made sense. Application documents were mixed with rejection letters, medical examinations, legal correspondence, and random bureaucratic forms. In the old days, I would have just read through all 173 pages, taking notes, trying to keep track of what connected to what. Now, I had a better strategy. I photographed key pages from the pension file. The ones that looked like official documents or affidavits. Then, I uploaded them to Claude in-sequence and asked it to create comparison table. Here's the exact prompt I used. Quote, I'm uploading several documents from a Civil War pension file. Please create a table comparing 1. Document type 2. Date 3. Who created it 4. What it claims to prove about the marriage 5. Strengths of the evidence 6. Potential weaknesses the pension office might see Focus specifically on whether each document meets the legal requirements for approving a valid marriage under Pennsylvania law in the 1890s. End quote. Now, this is what I call strategic document analysis. I'm not asking AI to tell me whether Catherine was married. I know she was. I'm asking it to help me see these documents the way the pension examiner would have seen them. Claude processed the documents and gave me this table that I'll upload into our Facebook group this week for you to see. And here's what Claude pointed out in its analysis. Something I hadn't thought of. None of these documents proved the marriage ceremony took place in 1858. They proved cohabitation. They proved a long-term relationship. They proved Samuel acknowledged Catherine as his wife and the children as his. But the pension office needed proof of the legal marriage itself. 

I sat with this for a while. Because here's the frustrating thing. In modern genealogy terms, this is rock-solid evidence. 

If I were writing a genealogical proof argument, I could make a strong case that Samuel and Catherine were married. But the 1892 pension office wasn't accepting genealogical proof arguments. They wanted specific documentation of the ceremony. Remember, Catherine mentioned in her letter that Pastor Mueller married us in 1858 and he is dead now, but his son remembers it. This gave me a thread to pull. If I could find information about Pastor Mueller, who he was, which church he served, whether his son really did remember the wedding, that might explain why Catherine's second application also got rejected. I went back to Perplexity and searched, German Reformed Churches, Berks County, Pennsylvania, 1850s, Pastors Church Records, end quote. Perplexity found several historical society publications and old county histories that mentioned German Reformed congregations in Berks County. There were at least six in the rural areas around Reading in the 1850s. But here's where I hit a problem. I didn't know which specific church it was. The pension file never names it. Catherine just says, Our Church. This is where using multiple AI tools strategically really pays off. I took the information Perplexity found about German Reformed churches and fed it into Claude along with Catherine's letter and the witness affidavits. I asked Claude, Quote, Based on the witness statements mentioning where the couple lived, near the old mill on Tulpehocken Creek, and the dates they give, which of these German Reformed churches was most likely their congregation, end quote. Claude analyzed the geographic references in the documents and suggested that based on the Creek location and the dates, it was probably either a manual Reformed church or Trinity Reformed church. Then I had an idea. In the pension file, there was a letter from the son of Pastor Mueller. It was written in response to a request from the pension office for verification of the marriage. But I'd skimmed over it before because the handwriting was nearly illegible. Worse than Catherine's. It was written by someone elderly with an unsteady hand. Now, knowing that Jim and I could handle challenging handwriting, I uploaded it to Google AI Studio. The transcription that came back was...heartbreaking. July 20th, 1892. To the Bureau of Pensions. You have written to ask if my father, Pastor Heinrich Mueller, performed a marriage between Samuel Hoffman and Katherine Schneider in 1858. I regret to tell you that I do not have clear memory of this wedding. I was but 14 years old at the time and my father performed many marriages. However, I can tell you this with certainty. The records of a manual reformed church were kept by my father with great care in a bound register. This register included all marriages from 1845 to 1889. When the church burned, the register was in the church office and was destroyed. I saw the register many times in my youth. It was written in German in my father's hand. If Samuel Hoffman and Katherine Schneider were married by my father, the record existed until the fire. I am sorry I cannot provide better testimony, but I can swear that if the marriage occurred at our church, the record once existed. Respectfully, Frederick Mueller. Do you see what happened? Frederick Mueller was trying to help. He was testifying that the records existed. He was explaining why they couldn't be produced. But from the pension office's perspective, this letter didn't prove the marriage. It just confirmed that the proof was gone. Now, before I draw any conclusions, I needed to verify what AI had helped me transcribe and analyze. And this is crucial. Remember our golden rule. AI is your research assistant, not your researcher. So, here's what I did. Number one, verified the churches. I looked up Emmanuel Reformed Church in Berks County historical records. It existed. It burned in 1889. All confirmed. Number two, verified Pastor Mueller. Found obituaries and church records confirming Heinrich Mueller was indeed the pastor there from 1840 to 1878. His son, Frederick, succeeded him briefly before the church burned. Number three, cross-reference the witnesses' names. Johann Schmidt and Elizabeth Weber both appear in the 1870 census near the Hoffman property. They were real people who really lived near the Hoffmans. Number four, checked a land deed. Ordered a copy from the Berks County courthouse. It matches what's in the pension file. Everything AI helped me transcribe and analyze? Accurate. The tools information. They helped me understand information that was already there. Using perplexity again, I researched how common this situation was. Quote, How many Civil War widow pension applications were rejected due to destroyed church records? What happened to these widows? End quote. And here's what I found. This was not unusual. Wooden churches burned regularly. County courthouses burned. Records were destroyed in floods, fires, and wars. And when that happened, widows were often left without proof. Some states passed special legislation allowing alternative evidence. Pennsylvania did eventually, but not until 1905. Thirteen years too late for Catherine. The pension office did sometimes accept common law marriage evidence, but it was inconsistent. Some examiners were more flexible than others. I need to pause here and talk about something we don't often address in genealogy research. For the information on this program. As I was going through these documents, reading Catherine's increasingly desperate letters, seeing Frederick Mueller's apologetic testimony, watching the pension office reject her application again and again, I started to feel the weight of this. This wasn't just research. This was someone's life falling apart in real time preserved in government documents. And this is where AI helps in an unexpected way. By making it easier to transcribe and understand these documents, AI helps us connect more deeply with our ancestors' lived experiences. We're not just seeing names and dates. We're reading their words. Understanding their struggles. Okay, so let me pull together everything I learned about Catherine Hoffman's situation. The basic facts. She was legally married to Samuel. By every standard that mattered. Cohabitation. Public acknowledgement. Children. Community recognition. They were husband and wife. Pennsylvania law recognized common law marriage. She should have qualified for the pension. But the pension office applied a stricter standard than the courts would have. They wanted documentary proof of the marriage ceremony itself. Or multiple affidavits from non-family members who witnessed the ceremony and could describe specific details. Catherine couldn't provide that because The church register burned in 1889. Two, the pastor who performed the ceremony died in 1878. Number three, most of the people who would have attended a small wedding in 1858 were dead by 1892. Number four, the pastor's son didn't have clear memory of one specific wedding from 34 years earlier. Her second application, the one from May 1892, was stronger than the first. She had neighbor affidavits, she had the midwife's testimony, she had official documents, but it still wasn't enough. Here's where the story gets more complex. And this is the part that remains partially unresolved. In October 1892, Catherine submitted a third application. And this one included something new, an affidavit from an elderly woman named Anna Schneider, who claimed to be Catherine's aunt. I uploaded this affidavit to Jim and I for transcription. The handwriting was incredibly shaky. This woman was clearly very old or very ill, but Jim and I got most of it. Anna stated that she was present at the wedding in 1858. She named Pastor Mueller. She gave specific details. It was a cold day in February. There were about 20 people present. Catherine wore her mother's dress. Samuel's brother stood up with him. This should have been the breakthrough. This was exactly what the pension office wanted. A witness who was at the wedding, who could provide specific details, who had no financial interest in the outcome. But, when I looked up the next document in the file, my heart sank. was a letter from the pension examiner noting that Anna Schneider had died in November 1892, six weeks after submitting her affidavit, and questioning the validity of the testimony since she couldn't be cross-examined. The pension file ends in December 1892 with a notation that says: Application remains under review. Additional evidence requested. But there are no more documents after that. No final approval. No final rejection. Just nothing. Now, I searched for Catherine in later records to see what happened to her. And here's what I found. The 1900 census lists Catherine Hoffman, widowed, living with her adult son, Jacob, and his family. She's listed as dependent, with no occupation. She's 58 years old. The 1910 census shows the same arrangement. She's now 68. Still living with Jacob. I found her death certificate. Catherine Hoffman died in 1916 at age 74. Cause of death: Pneumonia. She's buried next to Samuel in a small cemetery outside Reading. But here's what I didn't find. Any record that she ever received a pension. I checked the 1890s veterans schedule. She's not listed. I checked the 1883 pension roll. Not there either. As best I can determine, Catherine never got her widow's pension. She spent the last 24 years of her life dependent on her children. Living in what the census describes as their household. The farm they'd owned? The one Catherine mentioned in her desperate letter? I found it in later deed records. It was sold for unpaid taxes in 1893. Now I can't prove definitely that she was never approved. It's possible the file is incomplete. It's possible she was eventually approved and those records are somewhere else. But given what I've found, it seems most likely that she lost this fight. Here's what gets me about this story. Catherine did everything right. She gathered witness statements. She found official documents. She persisted through multiple rejections. She even got her elderly aunt to provide the exact testimony the pension office demanded. And it still wasn't enough. Or maybe it would have been enough if Anna hadn't died before the examiner could interview her. Or maybe not. The pension system, for all its good intentions, created bureaucratic hurdles that some widows simply couldn't clear. Not because they weren't entitled to the benefits, but because they couldn't produce documentation that didn't exist anymore. Or in some cases, had never existed in the first place. But here's what I learned from this research that applies to all of us doing genealogy. When you find a rejected application, whether it's a pension, a land claim, a citizenship petition, whatever, those rejections often tell you more about the bureaucratic system than about the actual truth of the situation. Catherine was married to Samuel. I'm as certain of that as I can be of anything in genealogy. The evidence, when you look at it holistically, is overwhelming. But the pension office couldn't look at it holistically. They had regulations. They had a checklist. And if you couldn't tick all the boxes, you were denied. Here's what I'm still trying to figure out. And this is where I need help from listeners who might have insights. In the pension file, there's a reference to a fourth application that Catherine supposedly submitted in 1893. The reference appears in a routing slip that says, quote, C-file C-4, additional application dated March 1893, end quote. But that file isn't in the pension records I received from the National Archives. It might be misfiled. It might be lost. Or it might have never been processed. If that fourth application exists, it could tell us whether Catherine ever succeeded. Or it might just be another chapter of this sad story. I've requested help from the archives to track down any additional files. But so far, no luck. Now, here's why this matters beyond Catherine's individual story. When I looked at her descendants, I found something interesting. Her son Jacob, the one she lived with after losing the farm, He became a county clerk. His daughter became a lawyer. His grandson became a judge. And I wonder if watching their grandmother fight this losing battle with the pension office influenced their choice of careers. Sometimes, our ancestors' struggles shape our families in ways we don't even realize. Alright, here's your challenge and homework for this week. Find a rejected application in your family tree. It could be a pension, a land claim, a citizenship petition, anything where your ancestor was told no. Then, use the techniques we covered today. 1. Research the historical requirements using perplexity. 2. Transcribe any handwritten documents using Gemini. 3. Analyze the evidence using Clod. 4. Verify everything with primary sources. Share what you find in our Facebook group, Ancestors and Algorithms, AI for Genealogy. I especially want to hear about rejected applications where your ancestors eventually succeeded. Because those stories need to be told too. And if anyone has experience researching Civil War pensions, Or, if you've ever found missing files at the National Archives, Please share your tips. I'm still looking for Catherine's fourth application, And I could use the wisdom of the community. I want to hear from you. you found pension files in your research? Have you discovered ancestors who fought bureaucratic battles like Catherine's? Email me at ancestorsandai at gmail.com Or, join the conversation in our Facebook group. Thank you so much for listening to Ancestors and Algorithms. If you enjoyed this episode, Please leave a review wherever you listen to podcasts. It helps other family historians find us. Remember, AI is your research assistant, Not your researcher. It helps us work faster and dig deeper. But it's our job to verify, To question, And to tell our ancestors' stories with accuracy and respect. I'm your host, Brian, And I'll see you next week For another journey into the past, Powered by the future. Until next time, Happy researching.