Day in the Life of an Architect: Three Real Days

Coffee's on. I'm standing at the kitchen counter scrolling through email before anyone else in the house is awake. Fourteen new messages overnight. Two from contractors, one from the city, and the rest are product rep newsletters I keep meaning to unsubscribe from. The one that matters is from Doyle, the general contractor on the Forrester house. He says the steel beam over the great room is in and he wants me out there this morning to look at it before they frame around it. I write back "9 AM works" and move on to the city email. Shonda in plan review says our zoning pre-application for the Delgado project is confirmed for 1 PM today. I forgot that was today. I pull up the site plan on my laptop to refresh my memory while I eat a bowl of cereal.

In the office. We rent the second floor of a converted house on South Congress. Four desks, a printer that jams every third job, and a long table covered in material samples and rolled-up drawings. Clementine is already here. She's been with me three years, right out of school, and she's the reason the Ortiz construction documents are actually going to be done by Friday. She's working on the reflected ceiling plan when I walk in. I ask how it's going. She says "fine, except the HVAC diffusers don't fit where we drew them." This is always the answer. The HVAC diffusers never fit where you draw them.

I spend 30 minutes reviewing Clementine's progress on the Ortiz set. The floor plans are solid. The wall sections need another round of dimensioning. The window schedule has a typo on the master bedroom units, which lists them at 4'-0" when they should be 4'-6". I mark it up on the screen and she fixes it in about ten seconds. These are the kinds of errors that seem tiny until a window shows up on a truck and it doesn't fit the rough opening. A six-inch difference means tearing out framing. I've seen it happen. It cost $3,400 and a week of delay on that project.

Driving to the Forrester site in Westlake. The Forresters are building a $1.2 million house on a slope that drops about 18 feet from the street to the back of the lot. We designed the house to step down the hill in three levels. The structural engineering on it was complicated. The foundation alone took six weeks. The framing started two weeks ago and Doyle's crew has been moving fast. I pull up to the site and the house is suddenly a real thing. Last time I was here, it was just foundation walls and dirt. Now there are walls and a roof structure. It always catches me off guard, that moment when a building goes from lines on paper to something you can walk through.

I'm standing in what will be the great room. Right now it's just studs and sky. The framing crew left for break so it's quiet except for a radio somewhere playing country music. I can smell fresh-cut lumber and that particular sweetness of pressure-treated wood baking in the sun. The steel beam Doyle emailed about is a W12x26, spanning 22 feet across the great room. I pull out my tape and laser level. The beam is supposed to sit at 9'-4" above finished floor. I measure from the top of the subfloor to the bottom of the beam. I get 9'-6". That's two inches higher than spec. Two inches doesn't sound like much. But it means the header above the sliding doors on the south wall, which ties into this beam, is also two inches high. Which means the door rough opening might be wrong. I photograph it, mark the measurement on the beam with a lumber crayon, and call Doyle.

Doyle meets me on site. He looks at the beam, looks at my measurement, and says "the steel guys set it where the structural drawings showed it." I pull up the structural engineer's drawing on my phone. He's right. The structural drawing shows the beam at 9'-6" to bottom of steel. My architectural drawing shows 9'-4" to bottom of steel. We have a two-inch conflict between the structural and architectural sets. This happens more than it should. I call Marcus, the structural engineer. He picks up on the second ring. After three minutes we figure out the discrepancy: his dimension is to bottom of beam flange, mine is to bottom of beam including the bearing plate. The bearing plate is two inches thick. We're both right. The beam is in the right spot. I feel equal parts relieved and annoyed that I spent 40 minutes worried about a dimension that was never actually wrong. I take a photo of the bearing plate for my records and write a note so this doesn't confuse anyone during framing inspection.

Back in the car. I spend 15 minutes in the parking lot of a gas station reviewing the Delgado site plan before this afternoon's meeting. The Delgado project is a duplex on a lot that's currently zoned single-family. The client wants to do an accessory dwelling unit under Austin's new ADU code, but the lot has a heritage tree near the back property line that limits where we can build. The pre-application meeting is a chance to run the plan past Shonda before we submit formally. She's been in plan review for 12 years and she knows the code better than most architects. When Shonda says something won't fly, it won't fly. I'd rather hear it now than after we've submitted and waited six weeks for a denial.

Back at the office. Clementine and I eat lunch at our desks. She's having some grain bowl thing. I'm having a turkey sandwich from the place next door. We talk about the Ortiz deadline. She thinks the set will be at about 90% by Thursday, which means I can do my review Thursday night and we submit Friday morning. That's tight. But tight is how we operate. Four people, three active projects, two in design and one in construction. We don't have the luxury of extra weeks.

I spend an hour on the Delgado setback calculations. The lot is 50 feet wide and 130 feet deep. Front setback is 25 feet. Side setbacks are 5 feet each. Rear setback is 10 feet. The heritage tree has a critical root zone radius of 18 feet from the trunk, which is 6 feet inside the rear setback. So our buildable area in the back is actually 18 feet from the tree, not 10 feet from the property line. I'm sketching options on trace paper over a site survey printout. I do this by hand because the feedback loop is faster than opening a CAD file. Pencil, trace paper, scale ruler. The same tools I used in school. The computer is for production. The thinking happens on paper.

City hall. The zoning pre-application meeting is in a conference room on the third floor. Shonda is already there with a stack of code books and a highlighter. I present the site plan. She listens, asks two questions, and then tells me the ADU setback from the heritage tree is actually measured from the drip line, not the trunk. The drip line extends 22 feet, not 18. I lose four more feet of buildable area. I ask if there's a variance process. She says yes, but it takes 90 days and requires a public hearing. I write that down. The Delgados are not going to love this. The meeting lasts 25 minutes. On the way out, Shonda mentions a recent code amendment that might let us reduce the rear setback to 5 feet for ADUs on lots over 7,000 square feet. The Delgado lot is 6,500. We're 500 square feet short. I write that down too.

Back at the office. I call the Delgados and walk them through what Shonda told me. They're disappointed but not surprised. They've been through one round of permit denial already on a previous design attempt with a different architect. I tell them I want to explore the variance route and also look at reconfiguring the ADU footprint to stay within the drip line setback. They say okay. Laura Delgado asks if this means it will cost more. I tell her the variance application is about $800 in city fees, plus my time. She asks me to think about it and send her options by end of week. I add it to my list, which is now longer than my week.

Clementine and I spend two hours on the Ortiz construction documents. She's drawing the kitchen cabinet elevations. I'm doing the exterior wall section detail at the roof parapet. This is the kind of work that doesn't sound like architecture to most people. It's not designing a building. It's drawing exactly how flashing wraps over a parapet cap, where the air barrier terminates, and what size screw holds the coping in place. But this is where buildings actually succeed or fail. A beautiful design with bad flashing details leaks in the first rainstorm. I've seen it. The building I'm most ashamed of is a house from my second year of practice where I didn't detail the deck-to-wall connection properly. Water got behind the siding within a year. The repair cost $11,000. The client was gracious about it. I think about that connection every time I draw a flashing detail, which is probably the appropriate penance.

Clementine leaves. I stay another 45 minutes working through the Ortiz plumbing coordination. The plumber needs a 3-inch waste line to drop through the floor in the upstairs bathroom, but there's a beam directly below where it needs to go. I could move the toilet 6 inches. I could ask the structural engineer to add a penetration through the beam. I could route the waste line horizontally to a different drop point. I sketch all three options and decide to ask the plumber which one he prefers, because he's the one who has to install it and his opinion on what's easiest is worth more than my opinion on what looks cleanest on paper.

Home. My wife Jenn asks how the site visit went. I tell her about the beam being two inches off. She says "is that a lot?" I say it turned out to be nothing, just a dimension measured from two different reference points. She says "so it was fine?" I say "it was fine but it took an hour to figure out it was fine." She nods the way she nods when I talk about architecture, which is polite but also clearly filing it under "Hale's work stuff that I don't need to retain." Fair enough. I wouldn't retain her day at the marketing agency either. We eat dinner. I think about the parapet detail. I think about the drip line. I think about the plumbing drop. The building keeps going in my head long after I leave the office. That's the part of this job that nobody warns you about. The drawings stop, but the thinking doesn't.

Biking to the office in Capitol Hill. It's 48 degrees and drizzling, which in Seattle is just called Thursday. My rain jacket is from REI and it cost more than my first week's take-home pay at this job. I lock up outside the building, a glass tower on Pine Street that our firm did not design. We're on the 11th floor. I take the elevator with two people I don't recognize. Our firm has 200 people across three floors and I've been here two years and still couldn't name half of them.

The office is half-lit. The motion sensor lights haven't triggered yet on the north side, so it's just the glow of a few early monitors and the coffee machine doing its thing. I can hear it gurgling from 40 feet away. Risa, my desk neighbor, isn't here yet. She started the same week I did and we've developed a routine where whoever gets in first makes the other person's coffee. Today that's me. Hers is oat milk, two sugars. Mine is black. I set both mugs on our shared shelf and open my laptop.

Revit is loading. This takes about four minutes because the model for the Lakeview project is 1.2 gigabytes. The Lakeview is a 14-story mixed-use building, retail on the ground floor, offices on 2 through 6, residential on 7 through 14. I'm working on the curtain wall system for the office floors. Specifically, I'm detailing the transition where the curtain wall meets a precast concrete spandrel panel at each floor line. This detail has to account for thermal expansion, waterproofing, structural movement, and aesthetics. It's about 8 inches of wall assembly and it's been the focus of my last three weeks.

Risa arrives. She sees the coffee. She says "bless you" the way she says it every morning, like it's a sneeze response. She's working on the residential floor plans, which are 147 units across 8 floors. She's been adjusting unit layouts to accommodate a structural column that shifted 6 inches in the latest structural coordination round. Six inches in a 550-square-foot studio apartment is not trivial. She's been moving kitchen counters and closet walls for days. We commiserate for about two minutes and then we both put our headphones on, which is the universal office signal for "I'm working now."

Deep in the curtain wall detail. I'm modeling the aluminum mullion profile, the structural silicone joint between the glass and the mullion, and the gasket that seals the system at the spandrel interface. The manufacturer's shop drawings show the gasket at 3/8 inch compressed. Our detail shows 1/2 inch. I email the curtain wall consultant to ask which dimension governs. While I wait for an answer, I model both versions so I can swap them when I get confirmation. This is how I spend a lot of my time: modeling two options because I'm waiting on an answer from someone who's busier than I am.

Meeting with the mechanical engineer about HVAC coordination on floors 2 through 6. The conference room has a whiteboard, a screen, and a table that seats twelve. There are four of us: me, Einar (my project architect), Wynn (the sustainability consultant), and Dan from the mechanical engineering firm. Dan's main concern is that our ceiling plenum is 14 inches and he needs 16 for the ductwork routing on the east side of the building. Two inches. Always two inches. Einar asks if the duct can be rerouted. Dan says maybe, but it pushes a supply diffuser into the curtain wall zone. I pull up my curtain wall section on the screen. We all stare at it. Wynn asks about the thermal bridging at the mullion. Einar writes something on the whiteboard. We agree that Dan will study the reroute and I'll check if the curtain wall head detail can accommodate a diffuser within 8 inches of the glass line. Meeting runs 40 minutes.

Back at my desk. The curtain wall consultant replied. The gasket is 3/8 inch compressed. I delete the 1/2-inch version from the model and update the detail sheet. I notice the section cut line on the floor plan doesn't align with the detail I just revised, so I fix that too. Then I notice that the keynote for the sealant joint references an old spec section. I update the keynote. This chain of small corrections is what documentation actually looks like. You fix one thing and discover three more things connected to it. It's like pulling a thread, except the thread is a 400-page drawing set and every thread connects to six others.

Lunch at my desk. Leftover pasta from last night. Risa is doing the same thing, eating a sad desk salad while staring at her residential floor plans. We talk for about five minutes about ARE exams. I'm studying for my third one, Project Development and Documentation. She just passed her second. The licensure process requires passing six division exams. I've been studying for about 14 months and I'm not halfway done. My firm pays for the exam fees, which is $235 per exam, but the study time is all mine. Evenings and weekends. Einar told me it took him three years to finish all six. He said the trick is to never stop, because if you take a break you lose momentum and the content starts to fade.

Einar stops by my desk and asks to see the curtain wall detail. I show him the updated section. He looks at it for about 30 seconds, says "the firesafing at the slab edge needs to be called out," and walks away. He's right. I forgot to add the firesafing detail where the curtain wall passes the floor slab. This is a code requirement, rated assembly, and I just... didn't draw it. I add it. It takes 20 minutes. In those 20 minutes I think about how a building envelope has about 200 details like this, each one a potential point of failure, and how the job of an architect is basically to remember all 200 of them before the building goes up. Einar remembers them. I'm learning to.

Back in the model. I'm now working on the curtain wall corner condition where two glass planes meet at 90 degrees. The corner mullion is a custom profile, wider than the typical mullion, and it has a different thermal break configuration. The manufacturer sent a DWG file of the profile. I import it into Revit as a family, which takes about 45 minutes because the DWG has about 30 extra layers of manufacturer annotation that I have to clean up before it's usable. I think about this sometimes: how much of my job is cleaning up other people's CAD files so they fit into our model. The answer is "more than you'd think."

Einar sends me a markup PDF of the floor plans with comments from the client's project manager. There are 23 comments. Most are questions I can answer. "What is the finish on the lobby columns?" Polished concrete. "Is the bike storage room ventilated?" Yes, there's an exhaust fan on Sheet M-204. A few require design decisions that Einar will have to make. I sort the comments into two lists: ones I can resolve and ones I need to escalate. I start working through my list. This is not what I pictured when I was in architecture school, sorting comment logs and cross-referencing spec sections. But it's what the job actually is at this stage.

The curtain wall consultant calls. She wants to discuss the corner condition I'm modeling. Her name is Petra and she's been doing curtain wall engineering for 15 years. She explains that the corner mullion I'm drawing won't work with the glass size we specified because the bite (the amount of glass captured by the frame) isn't deep enough to handle the wind load at the building's corner. She needs either a deeper bite or smaller glass panels. I take notes. This means the corner module width changes, which means the curtain wall layout on the east elevation shifts by about 3 inches, which means six sheets in the drawing set need updating. I tell Einar. He says "okay." That's his version of acknowledging that I just inherited two more days of work.

I start updating the east elevation to reflect the new corner module. Risa left at 5. The office is thinning out. The motion sensor lights on the south side have shut off. I can see the Olympics through the conference room window, just barely, through the clouds. I work through the elevation changes, updating dimensions, adjusting the curtain wall grid, moving keynote leaders. I save the model. The save takes almost two minutes because of the file size.

I close my laptop. Nine hours today, almost all of it in Revit, almost all of it on one section of wall. Tomorrow I'll keep going. The curtain wall package is due in three weeks and there are 14 detail sheets to finish. My work today will be one page, maybe one and a half, in a 400-page drawing set. I think about that on the bike ride home. One day of work, one page. The building will take two years to build and 50 years to stand there. There's something satisfying about that ratio, even if it doesn't feel like it when I'm cleaning up a manufacturer's DWG file at 4 in the afternoon.

Driving to the Calloway Mill. It's a 1920s textile mill about 25 minutes east of Nashville, a four-story brick building that's been empty since the late 1980s. A developer bought it last year and wants to turn it into a food hall on the ground floor with office space above. My job is to figure out what can change and what has to stay. The building is on the National Register of Historic Places, which means the renovation qualifies for federal historic tax credits, but only if the work meets the Secretary of the Interior's Standards for Rehabilitation. Those standards are 10 guidelines that boil down to: keep the character, don't fake the history, and make new work reversible where you can.

On site. The mill smells like old brick and machine oil, a hundred years of cotton dust and grease soaked into every surface. I love this smell. It's the smell of a building that actually did something. The light coming through the industrial windows is spectacular. They're steel sash, multi-pane, floor to ceiling on the south wall. Some are broken. Some are painted shut. Some have lost their glazing putty entirely and the glass is just sitting in the frame held by friction and optimism. I photograph each one. There are 47 windows on the south elevation alone.

Jules, my project coordinator, meets me on site with the measuring equipment. Jules has been with the firm for six years and he's the most organized person I've ever worked with. He carries a clipboard with pre-printed field forms, a laser measurer, a 100-foot tape, a moisture meter, and a mortar testing kit. I carry the camera and a sketchbook. We start at the northeast corner and work our way around the building. I'm documenting existing conditions: the size and spacing of every window, the condition of the brick, the state of the mortar joints, any visible cracking or settlement, the locations of later additions and modifications. Every detail matters because the preservation plan has to distinguish between original fabric and later alterations.

I'm testing mortar on the east wall. This is a hands-on process. I take a flat-head screwdriver and press it into a mortar joint at a 45-degree angle. If the mortar crumbles easily, it's lime-based, which is original to the 1920s construction. If it resists, it's Portland cement, which means someone repointed it later. The original lime mortar is softer than the brick, which is by design. It lets the wall flex slightly with temperature changes instead of cracking the brick. When someone repoints with hard Portland cement, the mortar becomes stronger than the brick, and the brick starts spalling. I find patches of Portland repointing on the northeast corner, probably from the 1970s based on the color. Three courses of brick above the repointing show classic spalling, chips and flakes where moisture trapped by the hard mortar froze and broke the brick face. I photograph it and write "inappropriate repointing, recommend removal and lime mortar replacement" in my notes.

Inside now. The ground floor is one open room, about 15,000 square feet, with a grid of cast-iron columns spaced 20 feet on center. The columns are beautiful. Fluted, with simple capitals, the kind of industrial detail that people in the 1920s put into a factory because they believed even a workplace should have dignity. The floor is hardwood, worn smooth by decades of machinery and foot traffic. The developer wants to keep the columns exposed and the floor refinished. Good. That's the right call. The food hall concept works with this space because the column grid creates natural zones for vendors. I'm already thinking about where the kitchen exhaust hoods will go and how to route ductwork without blocking the windows.

Alton, the contractor, arrives. He's been doing restoration work in Nashville for 20 years and he knows old buildings the way a mechanic knows engines, by sound and feel. He taps a brick on the west wall and says "hollow." We both know what that means: the brick has delaminated from the structural wythe behind it. He taps three more. Two are solid, one is hollow. I mark the hollow ones with blue painter's tape so we can map the extent of the delamination during the structural assessment. Alton asks about the window situation. I tell him 47 south-facing, probably 180 total, most need restoration. He takes a breath. Window restoration is expensive, about $800 to $1,200 per window depending on condition. At 180 windows, that's $144,000 to $216,000 just for windows. But replacement would jeopardize the tax credits. Restoration it is.

I spend 30 minutes on the second floor measuring the original stair. It's a dog-leg stair with cast-iron balusters and a wooden handrail. The handrail is maple, worn to a satin finish by 100 years of hands. The newel post at the landing has a chip in it. Someone bolted a fire extinguisher bracket into it, probably in the 1960s, and the bolt hole cracked the wood. I measure and photograph the damage. This stair can be restored. The balusters need paint removal and recoating. The handrail needs patching where the bolt was. The treads are worn but structurally sound. The stair doesn't meet current code for a commercial occupancy, the risers are 8.5 inches and code says 7 inches max, so we'll need a code variance or a second stair that does comply. I write "stair code variance: discuss with Genevra" in my notebook.

Lunch in the car. A sandwich from the gas station down the road. I sit in the parking lot of the mill and look at the building while I eat. The roofline has a stepped parapet with terra cotta coping. Two sections of the coping are cracked. I make a note. The loading dock on the west side has a steel canopy that's rusted through in two places. I make a note. A building like this generates 50 pages of notes in a single site visit. Every surface tells you something if you're looking.

I call Genevra at the State Historic Preservation Office. She's the reviewer who will evaluate our preservation plan and make the recommendation to the National Park Service for the tax credit certification. I've worked with her on four projects. She's thorough, fair, and completely unmoved by developer timelines. I describe the building and our approach. She asks about the windows immediately. I tell her we're planning full restoration. She says good. She asks about the floor. I say we're keeping the original hardwood and refinishing it. She says good. She asks about new mechanical systems. I say we're routing ductwork through the basement where possible and using the existing column grid to conceal vertical chases. She says "send me the preliminary plan and I'll give you informal comments before you submit the Part 2 application." This is the best possible response. Informal comments from Genevra save months of back-and-forth during formal review.

Back at the office. I start organizing the field photos. I took 214 photos this morning. Each one gets tagged with a location code (NE-1 for northeast elevation, first bay; GF-C4 for ground floor, column grid C4) and a brief description. Jules handles the measurement data. Between us we'll assemble an existing conditions report that documents every significant feature of the building: materials, dimensions, condition, alterations, and recommended treatment. This report becomes the foundation of the preservation plan, which becomes the basis for the tax credit application, which is worth about $2.8 million to the developer. So the report matters.

I start writing the preservation plan narrative. The first section is a statement of significance, explaining why this building matters historically. The Calloway Mill operated from 1923 to 1987. At its peak in the 1940s, it employed 340 people and produced cotton duck fabric for military tents. The building is significant under National Register Criterion A for its association with Nashville's industrial history and Criterion C for its architectural character as a representative example of early 20th-century mill construction. I write three paragraphs. Each sentence has to be defensible because Genevra will read it with a critical eye and the Park Service reviewer will read it with an even more critical one.

The developer, Craig, calls. He wants to know if we can add a rooftop deck for the office tenants. I tell him a rooftop addition is possible but it has to be set back from the parapet so it's not visible from the street. That's a Secretary of the Interior standard, new additions shouldn't alter the historic character of the building as seen from the public right-of-way. He asks how far back. I say I need to do a sight-line study, but probably at least 15 feet from the south parapet. He goes quiet. A 15-foot setback on a 90-foot-deep building means the usable rooftop area is small. He says "let's discuss it next week." Translation: he's going to push back on this. I prepare my arguments.

I finish the first draft of the statement of significance and save it. Jules sends me the completed measurement table for the south elevation windows. All 47, documented. Opening sizes, glass condition, hardware condition, operability, presence of storm windows. Twenty-three are operable. Eleven are painted shut. Thirteen have broken or missing glass. I scan the data. No two windows are in exactly the same condition. That's the thing about old buildings. They age unevenly. The south wall gets more sun and more rain than the north, so the south windows deteriorate faster. The east windows near the loading dock have more physical damage, probably from trucks backing into them. Each window is its own small story of use and neglect.

Driving home. I'm thinking about the mortar. About the difference between lime and Portland cement and how a well-intentioned repair in the 1970s caused damage that I'm now documenting in 2026. Preservation work is long-term thinking on a scale that most architecture doesn't touch. The decisions I make on this project will affect this building for the next 50 years. The building has already been here for 103. If we do this right, it'll be here for 103 more. That's a different kind of responsibility than designing something new.

Home. My daughter Lily, she's 11, asks what I did today. I say "I measured 47 windows." She says "that's a lot of windows." I say "it is." She asks if they were big windows. I say "floor to ceiling, built in 1923, and most of them still work." She thinks about this for a second and says "that's pretty cool, actually." It is pretty cool, actually. I don't always feel it in the moment, when I'm crouched in a stairwell measuring riser heights or scraping mortar with a screwdriver. But standing in that mill this morning, with the light coming through those hundred-year-old windows and the smell of a century of work soaked into the walls, I felt it. Not every day gives you that. But some days do.