For #97, currently on the uniform of Mario Edwards, Jr., I'll go with the previous holder of that number, Poona Ford.
When Ford was coming out of Texas as reigning Big-12 Defensive Lineman of the Year, I was frequenting the Seahawks fan site with the silly name. A guy there got me interested in Ford before the draft, and I was hoping the Seahawks would get him on Day 3 of the draft. He went undrafted and I was really happy when the Seahawks signed him as an undrafted free agent after the draft.
No team was willing to risk even a late seventh-round pick on a guy who had been chosen the Defensive Lineman of the Year for a major conference, and it was widely seen as being because Ford was "too short" to be an effective defensive tackle in the NFL. I was hopeful he might do well for two reasons. First, one of the reasons the basic height-and-weight rules of thumb exist for defensive linemen is because a smaller lineman can be easier to move, but I thought Ford's low center of gravity could actually help him, and at least compensate for him being shorter than most players at his position. Another reason taller linemen tend to be more successful is because taller linemen tend to have longer arms, so they tend to have better reach. But Ford has a wingspan like those of players 6'7" tall, so he's not at the same disadvantage as other sub-six-foot defensive tackles in terms of his reach on the initial "punch."
In the end, despite not having been drafted, Ford became the kind of player that can be a useful part of an NFL team's roster. He's going into his sixth pro season now, having been a starter for four of his five seasons so far. He's good enough to be a starting DT in the NFL, and he can obviously be a very useful part of a defensive-line rotation. That's tremendous value for an undrafted free agent. As I said when Ford became a regular starter in his second season, the Seahawks getting Ford as an undrafted free agent made 31 teams look stupid and one team look lucky. Any team that had risked a sixth or seventh-round draft pick on him would have been thrilled with the return on investment. Heck, his first four years would have been a pretty good return on a fourth-round pick.
Ford got me thinking about the idea that some of the "rules of thumb" used as drafting guidelines, like not taking relatively short defensive tackles, might be incorrect. I actually started a machine-learning project to see if maybe I could determine what factors that a team can know about a player as of the draft can actually help predict how well that player will do in the NFL. Because of the inspiration for the project and because I was making probabilistic projections (predicting a distribution of results, not a specific point result), I called the project Projection of Output Odds for NFL Aspirants (POONA). In the end, I absorbed POONA into a wider project to determine a player's performance in his next game (and over the rest of the current season) based on what's known up to that time about the player. Instead of a tool that could possibly be useful for a professional sports team, and therefore with limited possible users and a difficult path to monetization, I changed the focus of the project in a way that allows me to monetize it without needing the connections to be able to pitch it to an NFL team, and without depending on an NFL team being in just the right "place" (moment, maybe) to need what my system can do.
Poona Ford was the inspiration for a project, POONA, that evolved into multiple projects in multiple countries. So he's my #97.